BAILII is celebrating 24 years of free online access to the law! Would you consider making a contribution?
No donation is too small. If every visitor before 31 December gives just £1, it will have a significant impact on BAILII's ability to continue providing free access to the law.
Thank you very much for your support!
 |
[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback] | |
The Law Commission |
||
|
You are here: BAILII >> Databases >> The Law Commission >> Modernising Communications Offences A final report [2020] EWLC 399 (July 2021) URL: http://www.bailii.org/ew/other/EWLC/2021/LC399.html Cite as: [2020] EWLC 399 |
||
[New search] [Printable PDF version] [Help]
(Law Com No 399)
Modernising Communications
Offences
Presented to Parliament pursuant to section 3(2) of the Law Commissions Act 1965
Ordered by the House of Commons to be printed on 20th July 2021
HC 547
© Crown copyright 2021
This publication is licensed under the terms of the Open Government Licence v3.0 except where otherwise stated. To view this licence, visit nationalarchives.gov.uk/doc/open-government-licence/version/3 or write to the Information Policy Team, The National Archives, Kew, London TW9 4DU, or email: mpsi@,nationalarchives.gsi.gov.uk.
Where we have identified any third party copyright information you will need to obtain permission from the copyright holders concerned.
This publication is available at www.gov.uk/government/publications.
ISBN 978-1-5286-2771-9
CCS0621857840 07/21
Printed on paper containing 75% recycled fibre content minimum
Printed in the UK by the APS Group on behalf of the Controller of Her Majesty's Stationery Office
The Law Commission
The Law Commission was set up by the Law Commissions Act 1965 for the purpose of promoting the reform of the law.
The Law Commissioners are:
The Right Honourable Lord Justice Green, Chairman
Professor Nick Hopkins
Nicholas Paines QC
Professor Sarah Green
Professor Penney Lewis
The Chief Executive of the Law Commission is Phil Golding.
The Law Commission is located at 1st Floor, Tower, 52 Queen Anne's Gate, London SW1H 9AG.
The terms of this report were agreed on 30th June 2021.
The text of this report is available on the Law Commission's website at
Contents
Page
Evidence and problems of proof
Repeal of the existing communications offences
Consultation responses (repeal of existing offences)
Analysis (repeal of existing offences)
Consultation question 5 - definition of “harm”
Consultation question 4 - likely harm
Consultation question 3 - likely to cause harm to likely audience
Consultation question 6 - context and characteristics of audience
The Mental element - intention
Consultation question 8 - awareness of a risk of harm and intention
Consultation question 11 - reasonable excuse
Consultation question 12 - public interest
“Sent or posted” a communication
Consultation question 16 - extra-territorial application
CHAPTER 3: KNOWINGLY FALSE, PERSISTENT AND THREATENING
COMMUNICATIONS, AND FLASHING IMAGES
Section 127(2) of the Communications Act 2003
Part 1: Knowingly false communications
The proposed new offence: knowingly false communications
Part 3: Threatening communications
Victim impact and prevalence of threatening communications
Use of communications offences
Consultation question and response
CHAPTER 5: GROUP HARASSMENT, GLORIFICATION OF VIOLENCE
AND VIOLENT CRIME, BODY MODIFICATION CONTENT
Existing law: group harassment
Incitement or encouragement of group harassment
Knowing participation in “pile-on” harassment
Part 2: Glorification of violence and violent crime
Existing law, and the recommended harm-based offence
Justification and consultation responses
Part 3: Body modification content
Body modification: existing law
Consultation question and responses
The fault element - additional intent
Ancillary orders and special measures
Part 2: Consultee responses - glorification or encouragement of self-harm
Preliminary issue: online self-harm content and suicide “challenges”
Nonsuicide self-harm content and the harm-based offence
Encouragement of self-harm: should there be a specific offence?
Part 3: Encouragement or assistance of self-harm: a new offence
Practical mechanisms: prosecutorial discretion and enforcement
1.1 This report is the last of three Law Commission publications concerning harmful online communications and the role of the criminal law in combatting that behaviour.1 It is in this report that we make our final recommendations for law reform. The recommendations in this report are designed to modernise the criminal law: to ensure the law works effectively in the new technological paradigm, and is future-proofed; to ensure the criminal law can be a useful tool in protecting people from genuine harm and abuse; and, in a world where we communicate widely and publicly to share or challenge complex and controversial ideas, to ensure the criminal law allows space for that discussion and does not disproportionately interfere in people’s legitimate freedom of expression.
1.2 We should be clear that the scope of these reforms extends beyond online communications. While it is in the online space that the limitations of the existing law have been most illuminated, it does not follow that reforms should focus solely on online communications. Indeed, the overwhelming tenor of the evidence we have seen is that the specific form of technology used is largely irrelevant to the potential for harm.
1.3 How we communicate - in mode, in frequency, in content - has fundamentally changed over the last two decades. It is worth recalling the opening paragraph of our consultation paper:2
When we published our report in 19853 leading to the enactment of the Malicious Communications Act 1988, only 12.6% of UK households owned a home computer - and the internet wouldn’t be invented for another four years.4 Even in 2003, when the Communications Act 2003 was passed, only 46% of UK households had internet access.5 It was another year before Facebook was released (and, even then, available only to select universities and colleges). Twitter was released in 2006. The first iPhone was not released until 2007, four years after enactment of the Communications Act. Nearly two decades later, in every single second that passes,
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381; Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 1.1.
Law Commission, Poison-Pen Letters (1985) Law Com No 147.
J Schmitt and J Wadsworth, ‘Give PC’s a Chance: Personal Computer Ownership and the Digital Divide in the United States and Great Britain’ (2002) Centre for Economic Performance, London School of Economics, p 17.
Office for National Statistics, Internet Access - households and individuals, Great Britain: 2018 (2018), p 3. nearly three million emails are sent, 1nearly ten thousand Tweets are posted, 2and over three quarters of a million WhatsApp messages are sent [which by late 2020 was actually 1.15 million per second],3 amongst the other c.95 terabytes of internet traffic per second.4
1.4 The suite of criminal offences concerned with communication has, by comparison, changed very little. The constituent elements of section 127 of the Communications Act 2003, for example, (which prohibits certain communications sent over a public electronic communications network), draws heavily on its ancestor, section 10(2)(a) of the Post Office (Amendment) Act 1935. The section 127(1) offence makes it an offence for a person to send a communication over a public electronic communications network (which would include the internet but exclude, say, Bluetooth or an intranet) that is grossly offensive, indecent, obscene or menacing. The Malicious Communications Act 1988 prohibits communications that contain threats or are grossly offensive or indecent; whilst it does not restrict the scope of the offence to communications sent over a public electronic communications network, the communication does have to have been sent “to another”, so limiting the use of the offence to address communications posted in public fora such as Twitter and other social media platforms and websites. It is perhaps no surprise that criminal laws predating widespread internet and mobile telephone use (to say nothing of social media) are now of inconsistent application.
1.5 The various challenges these formulations of offences pose can be generalised in two ways. First, the laws under-criminalise. Many culpable and damaging communications are left without appropriate criminal sanction. The current offences prohibit particular categories of content and, in some cases, particular modes of communication, but some abusive, stalking and bullying behaviours, despite causing substantial harm, simply fall through the cracks.
1.6 Secondly, at the same time, the offences over-criminalise. The scope of the existing offences permits prosecutions that, absent the careful prosecutorial guidance we have seen, would be so great in number as to swamp the criminal justice system, and may nonetheless constitute an unjustifiable interference in freedom of expression. By proscribing content on the basis of apparently universal standards - such as “indecent” or “grossly offensive” content - the law criminalises without regard to the potential for harm in a given context. Two consenting adults exchanging sexual text messages are committing a criminal offence, as would be the person saving sexual photographs of themselves to a “cloud” drive.
1.7 An even more significant problem is that the categories - particularly, though not uniquely, gross offensiveness - lack anything like a universally accepted definition. As we discuss in chapter 2, it will be difficult for a jury or magistrate to define whether something is offensive as a matter of fact without resorting to what they themselves consider to be offensive (let alone whether it was grossly offensive). As a result of this, it is very difficult to mediate conflicting opinion. One person might find a message to be profoundly offensive that another does not regard as offensive at all. It is not clear by what set of facts we might say that one person is right and the other wrong (and, by extension, to know whether the message was “grossly offensive” in law or not). The categories are therefore inherently susceptible to subjective definition, in that the criminal standard risks being derived from a jury or magistrates’ own opinion or experience.
1.8 That the existing criminal offences are inadequate to the task of protecting people from harmful communications has been a leitmotiv running both through our publications on the matter and through the numerous responses we received to our public consultation. That consultation, which followed the publication of our consultation paper in September 2020 and ran until December 2020, prompted many thoughtful and helpful responses; we are grateful for these and for the evidence therein. The need for reform was reflected in nearly all of the consultation responses: these came not only from across government and the criminal justice system, but also from independent groups such as victims’ charitable organisations and free speech groups.
1.9 The nature and extent of online harms are canvassed extensively in our scoping report5 and consultation paper. 6As we noted in the consultation paper,7 the online and offline worlds cannot neatly be separated: online hostility often maps “in-person” hostility. However, some forms of harmful communication are particularly prevalent in the online environment. This is for a variety of reasons, including the fact they are facilitated by online communications technologies, and in some instances fostered by the characteristics and cultural norms of the internet. 8While we do not propose to recapitulate the analysis we set out in detail in the scoping report and consultation paper, we provide the following overview by way of context for our recommendations in the following Chapters.
1.10 The scale and varieties of online harm were the subject of a number of studies that we cited in the consultation paper:9
(1) In a 2019 report commissioned by Ofcom, Jigsaw Research found that 61% of the 2,057 surveyed adults10 had at least one potentially harmful experience
online in the last 12 months.11 Forty-seven per cent had experienced potentially harmful content or interactions with others, and 29% per cent had experienced something they rated as “harmful” (annoying, upsetting, or frustrating).12
(2) In a 2019 report, the NSPCC found that 31% of the 2,004 children 13they surveyed had seen worrying or nasty online content, and 19% of surveyed secondary school students had seen content which encouraged people to hurt themselves.14
1.11 The prevalence of online abuse has sadly intensified in the context of the COVID-19 pandemic. A September 2020 report by the charities Fix the Glitch and the End Violence Against Women Coalition noted that “...the COVID-19 pandemic and associated lockdowns [created] a breeding ground for online abuse and [posed] new challenges for governments, employers, private companies and broader society.”15
1.12 As the Alan Turing Institute set out, COVID-19 has had a range of effects on online hate:16
COVID-19 has not only inflicted massive health costs, it has also amplified myriad social hazards, from online grooming to gambling addiction and domestic abuse. In particular, the United Nations High Commissioner for Human Rights has warned that the pandemic may drive more discrimination, calling for nations to combat all forms of prejudice and drawing attention to the ‘tsunami of hate’ that has emerged.
1.13 High profile figures are particularly susceptible to online abuse. According to data collected by the National Policing Advisor for Hate Crime, Paul Giannasi, and the National Online Hate Crime Hub, the young climate change activist, Greta Thunberg, is subject to, on average, 151 hateful posts per minute worldwide. When she met with former US President, Barack Obama, this rose to over 500 posts per minute.17 Another high-profile example of the pervasive and corrosive nature of online abuse is the ongoing abuse of Premier League football players. This has included widespread racist abuse of players and led to a boycott of social media platforms by the Premier League in response.18
1.14 The above merely provides a sample of the various instances of online abuse. We also repeat our observation that, whilst much of the focus is on online or digital communications, we have consistently worked to ensure (as much as possible within our terms of reference) our recommendations are technologically neutral. 19As we set out in the relevant sections of the report that follow, consultees have outlined their own experiences in their responses. We are extremely grateful for the generosity that all consultees and stakeholders showed in responding so extensively to our consultation, particularly in being willing to share their own, often harrowing, experiences of online abuse.
1.15 We analysed Article 10 of the European Convention on Human Rights (“ECHR”), which protects freedom of expression, at length in Chapter 2 of the consultation paper. For present purposes, it suffices to recall that the right to freedom of expression is a qualified right, in the sense that it can be subject to legitimate interference, but interferences in that right have to be justified. They have to be sufficiently prescribed by law, in pursuit of a legitimate aim, as well necessary in a democratic society.
1.16 As with any criminal offence concerning expression - which would include, say, breaches of the Official Secrets Act 1989, or fraud by misrepresentation under section 2 of the Fraud Act 2006 - the offences that we recommend constitute an interference in freedom of expression. The question, though, is whether the reason for the interference - the “aim” - is legitimate and whether the interference goes further than is necessary in a democratic society (which imports an assessment of whether it is proportionate).
1.17 As we note below, and is the focus of Chapter 2 of this report, the main offence that we recommend is based on the potential for harm; it is directly focused on the legitimate aim of protecting the health and rights of others. By focussing on communications that are intended to - and likely to - cause harm, rather than focussing (as the current law does) on broad categories that may or may not track harm, we necessarily have an offence that is context-specific rather than one based on general or universal standards such as “gross offensiveness”. Many consultees were concerned about the vagueness that, they argue, might be introduced by a move towards context-specificity, and the implications this would have for freedom of expression. As we noted in Chapter 2 of the consultation paper, we do recognise the importance of a sufficiently clear offence - that it is “adequately prescribed by law”. However, we think the concern is misplaced as it relates to our recommended offence:
(1) First, it assumes a certainty in the current law that we submit is elusive: the proscribed categories might hint at universal definitions, but this masks their intense subjectivity.
(2) Secondly, the defendant must have intended harm; this is not a law so vague that the innocent unwary may stumble within its scope.
(3) Thirdly, context-specificity is the solution to the problems associated with
universal standards; there is a sense in which the rejection of universal standards necessitates context-specificity. For example, an intimate image may be entirely harmless in one context, but clearly very harmful in another (using that image as a way of threatening someone, for example, is clearly very different from partners sharing intimate images of each other consensually between themselves, yet the content of the communication may be identical in each case). A similar point may be made of two friends sharing jokes between each other that they find amusing that, in another context, may be very distressing. So, we might not be able to say as a generality that a particular form of words or image will always be criminal or never criminal (because they either are or are not, say, “indecent” or “offensive”), but that doesn’t mean that we won’t be able to discern on a given set of specific facts that harm was or was not likely (and, of course, that is rather the point of the form of our recommended offence: we should only criminalise those communications with the real potential for harm).
1.18 In this report, we recommend a number of criminal law reforms designed to ensure that the law addresses genuinely harmful behaviour, whatever the form of technology used, and better protects freedom of expression.
1.19 This project has been funded by the Department for Digital, Culture, Media & Sport (“DCMS”) as part of the Government’s Online Harms strategy. At the time of writing, the Government intends to introduce legislation that would establish a new statutory duty of care. Compliance with this duty of care will be overseen and enforced by OFCOM. The regulatory framework will apply to search engines and “user-to-user services”, defined in clause 2(1) of the Draft Online Safety Bill 2021 as
internet services by means of which content that is generated by a user of the service, or uploaded to or shared on the service by a user of the service, may be encountered by another user, or other users, of the service.20
1.20 This regulatory framework is important context for our reforms, but it is not the focus of those reforms. This report is concerned to address the criminal law provisions that apply to individuals and not the liability of platforms. The inescapable reality is that the criminal law is a relatively cumbersome and expensive tool: it should be, and can only be, reserved for more seriously culpable communications. It is an important part of the solution to the problem of harmful communications, but a necessarily limited one.
1.21 The following terms were agreed with DCMS.
(1) The overall purpose of this project is to recommend reform of the criminal law relating to abusive and offensive online communications, following a period of public consultation.
(2) The project will cover:
(a) Reform and potential rationalisation of the current communications offences as they relate to online communication, with the aim of achieving a coherent set of offences that ensure online users are adequately protected by the criminal law.
(b) Where necessary, consideration of the meaning of “obscene”, “grossly offensive”, and “indecent”, as well as the definitions of “publish”, “display”, “possession”, and “public place” and of how we can best achieve clarity and certainty of the law for people communicating online.
(c) Consideration of online communications which amount to the glorification of self-harm and of violent crime and how these might be addressed in the context of any reform of the communication offences.
(3) In addressing the topics above, the Commission will be mindful to ensure that any recommendations for reform do not create incoherence or gaps in the criminal law as it relates to abusive and offensive communication offline.
(4) The Law Commission will also conduct a review to consider whether coordinated harassment by groups of people online could be more effectively addressed by the criminal law. This work on group harassment will form an extension of the online communications project.
1.22 The following areas are expressly outside the scope of this project:
(1) terrorist offences committed online;
(2) child sexual exploitation; and
(3) platform liability.
1.23 Our recommendations therefore do not address these topics.
1.24 The Law Commission is also working on two separate but related projects, namely Hate Crime and Intimate Image Abuse. 21Offences that fall within the scope of intimate image abuse (where intimate images of a person are taken, made or shared without their consent) may well also fall within the scope of the harm-based communications offence that we recommend. However, some instances of intimate image abuse are sufficiently serious that the level of culpability cannot adequately be captured by a communications offence. Further, a conviction for a communications offence will not carry with it the range of possible ancillary orders (such as Sexual Harm Prevention Orders).
1.25 Also, as we noted in the consultation paper, it is worth specifically recalling that the Hate Crime project and this project deal with different (albeit occasionally overlapping) issues. A proportion of online abuse can be, and often is, described as “online hate”. Indeed, a significant subset of online abuse is targeted at people on the basis of their race, religion, sexual orientation, transgender status, or disability - the characteristics currently protected by hate crime law. However, not all abusive online communications amount to online hate. Equally, hate crime can encompass a wide range of behaviour - including, for example, acts of physical violence against people because of their race or sexual orientation, or criminal damage to businesses or places of worship - as well as hate speech.
1.26 Our consultation paper was published in September 2020. There followed a three-and-a-half-month public consultation on the 30 questions that we had posed to consultees, including questions relating to our proposals for new criminal offences.
1.27 We received 132 written responses and conducted over 30 meetings and round-table events with stakeholders during that consultation period. Many consultees provided lengthy, considered, and well-evidenced responses. On the whole, consultees were supportive of both the rationale for reform and of the proposals we made. In a couple of instances, notably cyberflashing and the encouragement of self-harm, consultees’ responses caused us to reconsider the position that we had adopted in the consultation paper. We are grateful to them.
1.28 Unfortunately, some of our proposals were conflated misleadingly with the proposals made in the Hate Crime project and, in a number of instances, our policy position was presented as the precise opposite of the position we had actually adopted in the consultation paper. It is thus worth noting that a number of consultation responses were responding to proposals that we had not made. One area of widespread misunderstanding related to the “lack of reasonable excuse” element of the harmbased communications offence (noted below and discussed in detail in Chapter 2): many people thought that we had proposed this operate as a defence (ie the defence would have a burden of proving that the defendant had a reasonable excuse), when in fact it was always our intention that the prosecution would have to prove that the defendant lacked a reasonable excuse (so, in effect, the presumption is that a communication was reasonable). There was also some concern voiced to us that we had proposed criminalising certain forms of political speech merely because some people would be distressed by that speech; again, this was not our intention, and we even stated so expressly in the consultation paper. 22We hope that this report will address some of these misconceptions and the concerns arising therefrom.
1.29 Finally, we received a particularly helpful consultation response from the Northern Ireland Department of Justice. Our Terms of Reference, and statutory remit, are limited to the law of England and Wales. However, we were grateful for their perspective on our provisional proposals. They noted that telecommunications legislation is a reserved matter, 23but that Judge Desmond Marrinan’s report on Hate Crime legislation in Northern Ireland 24was well placed to consider the issue and assist Northern Ireland’s contribution to the wider debate.25
1.30 In this report, we recommend the following new or reformed criminal offences:
(1) a new “harm-based” communications offence to replace the offences within section 127(1) of the Communications Act 2003 (“CA 2003”) and the Malicious Communications Act 1988 (“MCA 1988”);
(2) a new offence of encouraging or assisting serious self-harm;
(3) a new offence of cyberflashing; and,
(4) new offences of sending knowingly false, persistent or threatening communications, to replace section 127(2) of the CA 2003.
1.31 One of the matters that we considered at length was the issue of flashing images being sent maliciously to known sufferers of epilepsy. This is the only instance that was brought to our attention of a physical harm arising directly (rather than consequentially, such as self-harm) upon seeing a communication. There is currently no specific offence of sending flashing images in our criminal law, and we could not guarantee that all instances of the behaviour could be prosecuted under the Offences Against the Person Act 1861. We did not consult on a specific form of criminal offence for this behaviour, and therefore make no recommendations as to the form of the offence, but we do recommend that the government considers introducing a specific offence to address this reprehensible behaviour.
1.32 We also consider the growing phenomenon of pile-on harassment, as well as the glorification of violent crime. However, as we explain, we do not make recommendations for reform. We were not persuaded that specific offences to tackle these matters would be necessary, effective, or desirable. In the case of pile-on harassment, it is our view that the criminal justice system is not the appropriate means of addressing the problem: the scale is simply not amenable to a criminal law solution. In any case, the existing criminal law provisions (albeit perhaps under-utilised) are able to deal with the more serious aspects of this problem (such as co-ordinated harassment). In the case of the glorification of violent crime, we were not persuaded that a criminal offence would be a proportionate or appropriate response (save where a specific violent crime was incited, though that would already be a criminal offence).
1.33 The offence that we recommend should replace section 127(1) of the CA 2003 and the MCA 1988 is an offence based on a communication’s potential for harm rather than on its content fitting within a proscribed category (such as “grossly offensive” or “indecent” content). Specifically, we recommend that it should be an offence for a person to send or post a communication (letter, electronic communication, or “article”, in the sense of “object”) that is likely to cause harm to a likely audience, intending that harm be caused to that likely audience. “Harm” for this purpose is defined as psychological harm amounting to at least serious distress.
1.34 This moves the focus away from broad categories of wrongful content, as we see in the current offences, to a more context-specific analysis: given those who were likely to see the communication, was harm likely or not likely? This formulation therefore ensures that, first, communications that are genuinely harmful do not escape criminal sanction merely because they cannot fit within one of the proscribed categories. Secondly, communications that lack the potential for harm are not criminalised merely because they might be described as grossly offensive or indecent etc.
1.35 We should also note that the threshold of harm - serious distress - is a higher threshold than under the existing law (indeed, neither the section 127(1) CA 2003 offence nor the MCA 1988 offence has any threshold of “harm”). Our concern is that the categories in the existing offences bring within scope too many communications of relatively low culpability. This was an especial concern bearing in mind the impact of the criminal law on the young and the vulnerable.
1.36 The fault element for our recommended offence is set at a higher level of culpability than is the case under the current law: the defendant must intend to cause serious distress. Under section 127(1) CA 2003, the defendant need only have intended to have sent the communication of the proscribed character (ie they need not have intended any particular result). The MCA 1988 offence requires proof that the defendant intended to cause alarm or distress, which is the same fault element as harassment (but, notably, harassment requires a course of conduct, whereas the communications offence is complete at the point of sending a single communication). Our threshold is higher; it requires proof that the defendant intended to cause psychological harm amounting at least to serious distress.
1.37 We also note that this offence differs slightly from the one provisionally proposed in our consultation paper. The original offence could be made out if the prosecution proved that the defendant intended or was “aware of the risk” of causing harm. We no longer consider that awareness of a risk constitutes an appropriate threshold of culpability for the purposes of this criminal offence.
1.38 We recommend that a new offence be included in the Sexual Offences Act 2003 to address a harmful and increasing phenomenon, popularly termed “cyberflashing”. Depending on the form of offending behaviour, this is a problem that is either not addressed well or not addressed at all by the current law.
1.39 The act of cyberflashing generally involves a person sending an unsolicited image of genitalia to another. It is important to distinguish cyberflashing from other forms of intimate image abuse where the victim is the subject of the image; in cyberflashing, the victim is not the subject but the recipient.
1.40 In our consultation paper, we proposed that the problem of cyberflashing be addressed by amending section 66 of the Sexual Offences Act 2003 (exposure) to include the sending of images or video recordings of one’s genitals. However, with the benefit of evidence provided in consultation, we are now of the view that this solution is too narrow in scope and is susceptible to problems of proof. We recommend that it should be an offence for a person to send an image or video recording of genitals (whether the sender’s or not) to another, either intending to cause that person alarm, distress or humiliation, or, where the image was sent for a sexual purpose, reckless as to whether it would cause alarm, distress or humiliation.
1.41 We remain of the view that the offence should be a sexual offence, and specifically one that allows for ancillary orders such as Sexual Harm Prevention Orders and Notification requirements, colloquially known as requirements to sign the sexual offenders’ register.
1.42 We make a recommendation for an offence of encouraging or assisting acts of serious self-harm. (This is similar in form to the offence of encouraging or assisting suicide in the Suicide Act 1961).
1.43 In our consultation paper, in accordance with our terms of reference, we asked consultees about an offence of “glorification” of self-harm. We were of the view that an offence based on glorification would be too broad, but that there may be scope for an offence targeting the most serious acts of encouragement.
1.44 We are now confident that, in serious cases, the criminal law can offer protection to victims of self-harm. However, being an area of complex public health policy, this is a difficult area in which to introduce criminal offences: the risk of criminalising vulnerable or young people is acute. We therefore recommend a narrow offence based on the intentional encouragement or assistance of self-harm amounting to grievous bodily harm (whether or not that harm resulted).
1.45 Finally, we recommend that section 127(2) of the CA 2003 be replaced with three separate offences addressing knowingly false harmful communications, hoax calls to public services, and threatening communications.
1.46 In our consultation paper, we asked consultees about each possible offence. We provisionally proposed the offences addressing knowingly false harmful communications and hoax calls to public services and asked for consultees’ views on an offence addressing threatening communications.
1.47 Consultees expressed extremely strong support for each of the offences. We are confident that our provisionally proposed offences targeting harmful false communications and hoax calls to public services are appropriate. Reflecting on the strength of consultees’ views and responses regarding threatening communications, we also recommend a specific offence to address them.
1.48 Though the scope of our project extends only to the substantive criminal law (ie criminal offences), it is nonetheless difficult to ignore the evidential challenges that the internet poses, especially given the resources available for investigating and prosecuting relatively low-level offences. The effect of virtual private networks (“VPNs”) - which, though increasingly popular for privacy and security reasons, allow people to hide their location and so to operate anonymously online - and anonymous accounts online means that effective investigation and prosecution can take many months, if it is possible at all.
1.49 We do not, and cannot, make any recommendations with respect to anonymity and VPNs. It is not within our terms of reference. What should be private online, and the extent to which internet users should be required to prove their identity, are questions with ramifications that reach far beyond the sphere of criminal law - and certainly beyond the scope of a review of the communications offences. The question of anonymity online therefore remains a matter for Government.
1.50 In the consultation paper we discussed the impact that online abuse has on children and young adults. 26While the experience of children and young adults differs according to their varying identities and characteristics, there are some specific points that can be made about the experience of children and young adults.
1.51 As we set out in the consultation paper:27
The Alan Turing Institute found that young people are more likely to experience abuse. Specifically, they found that:
41.2% of 18-30 year olds had seen cruel/hateful content online, compared with 7.4% of 76+ year olds. Age also impacted whether respondents had received obscene/abusive emails but the relationship was far weaker, ranging only from 13.1% of 18-30 year olds to 6.77% of 76+ year olds.
We have heard from some stakeholders that, amongst young people - especially University students - there is a culture of normalisation or dismissiveness about the harmful impacts of online abuse, with both perpetrators and victims sometimes using the excuse of ‘banter’ or ‘humour’ to diffuse the seriousness of this behaviour.
However, we also heard that this is consistent with an awareness of a risk of harm: the perpetrator may realise the seriousness of their actions, but attempt to minimise this by using humour as an excuse.
Moreover, normalisation or dismissiveness can exacerbate the harmful impact of online abuse. It may lead to inappropriate strategies being adopted or recommended - that victims should come off social media, for example. Moreover, it may cause victims to supress their emotional response, potentially contributing to more serious psychological harms later down the line.
1.52 A number of consultees, including the Magistrates’ Association and the Law Society of England and Wales, raised the issue of how our proposals would impact children and young adults. While they recognised that it is important to ensure that there are flexible and appropriate ways to sanction serious forms of online abuse, they argued that any new criminal offences should not disproportionately impact children and young adults.
1.53 The ubiquity of technology means that the potential opportunities for children and young adults to communicate have expanded exponentially since the existing communications offences were enacted. As set out in the Crime and Disorder Act 1998, the aim of the youth justice system is to prevent offending by children and young persons. 28Further, as we noted in our consultation paper that examined the issues of taking, making and sharing intimate images, the more nuanced approach that the criminal justice system takes to child offenders reflects the fact they are both still developing and underdeveloped in terms of their ability to understand the consequences of their actions.29
1.54 Whilst we consider the impact of our recommendations on children and young adults in a number of specific contexts throughout the report, 30two aspects of the potential impact of our recommendations on children and young adults are particularly important and we consider them here: charging and sentencing.
1.55 The Crown Prosecution Service guidance on youth offenders examines the issues that attend charging decisions and prosecution of youth. 31Further, in the context of the existing communications offences, the Crown Prosecution Service guidelines on prosecutions involving communications sent via social media are helpful in appreciating how an offender’s maturity (or lack thereof) is taken into account. They set out the factors that ought to be considered when deciding whether a prosecution is in the public interest, including the maturity of perpetrators:32
30. Prosecutors must be satisfied that a prosecution is required in the public interest and, where Article 10 is engaged, this means on the facts and merits of the particular case that it has convincingly been established that a prosecution is necessary and proportionate. Particular care must be taken where a criminal sanction is contemplated for the way in which a person has expressed themselves on social media.
31. Prosecutors therefore should, where relevant, have particular regard to:
a. The likelihood of re-offending. The spectrum ranges from a suspect making a one-off remark to a suspect engaged in a sustained campaign against a victim;
b. The suspect’s age or maturity. This may be highly relevant where a young or immature person has not fully appreciated what they wrote;
c. The circumstances of and the harm caused to the victim, including whether they were serving the public, whether this was part of a coordinated attack (“virtual mobbing”), whether they were targeted because they reported a separate criminal offence, whether they were contacted by a person convicted of a crime against them, their friends or family;
d. Whether the suspect has expressed genuine remorse;
e. Whether swift and effective action has been taken by the suspect and/or others for example, service providers, to remove the communication in question or otherwise block access to it;
f. Whether the communication was or was not intended for a wide audience, or whether that was an obvious consequence of sending the communication; particularly where the intended audience did not include the victim or target of the communication in question.
g. Whether the offence constitutes a hate crime (which may mean Article 10 is not engaged, but may also be a factor tending in favour of a prosecution in the public interest).
1.56 In relation to sentencing, the Sentencing Council’s guidelines on sentencing children and young people (which apply only to those under 18 years of age) explain the emphasis on prevention of offending and welfare of children and young people more broadly. These set out that, amongst the various considerations brought to bear, regard is had to a young person’s age and maturity, the seriousness of the offence, their family circumstances and any previous offending history. 33In our view these considerations could be of particular importance if considering offending that arose in the context of online communication.
1.57 In sum, we are of the view that the specialist approaches outlined above to both charging and sentencing apply to our recommended offences. We agree that, as set out by the Magistrates’ Association and Law Society of England and Wales in their responses to our consultation, there should be an emphasis on early intervention and education in respect of offending by children and young adults. We are also of the view that the various initiatives being undertaken as part of the wider Government response to Online Harms are relevant to this, including improving online media literacy and awareness for children, young people and adults.34
1.58 As well as having received 132 written responses to the consultation, many individuals and organisations gave their time to meet with us and provided valuable insight. As a result of the pandemic, many of these people have been under exceptional pressure yet chose to assist nonetheless. We are grateful to all those with whom we have met during the course of this project, for your time and for your expertise.
1.59 In particular, for their time as we prepared this report, we should like to extend thanks to the Crown Prosecution Service, the National Crime Agency, Samaritans, the Premier League, the Internet Association, Twitter, Google, Facebook, Zoom, the Alan Turing Institute, South West Grid for Learning, Professor Clare McGlynn QC (Hon), Dr Kelly Johnson, Sophie Gallagher, Mr Justice Martin Spencer, Professor Jacob Rowbottom, Professor Alisdair Gillespie, Dr Laura Scaife, The Law Society of England & Wales, English PEN, the Free Speech Union, Ofcom, the National Police Chiefs’ Council, the Magistrates’ Association, the News Media Association, and the Bar Council of England & Wales.
1.60 We are grateful to the following members of the Law Commission who worked on this report: Dr Nicholas Hoggard (lead lawyer); William Scobie (research assistant); and David Connolly (team manager).
2.1 In this chapter, we present a new communications offence that we recommend should replace the existing communications offences in section 127(1) of the Communications Act 2003 and the Malicious Communications Act 1988 (the “MCA 1988” and the “CA 2003”, and which, for the purposes of this chapter, are collectively termed the “communications offences”). These offences are not adequate to the task of criminalising communications in the twenty first century, particularly given fundamental changes in how individuals communicate. We consider that this new offence is crucial to ensure that the criminal law can be used as an effective tool against harmful communications - particularly online - while better protecting freedom of expression.
2.2 At first blush, it might appear odd to suggest that a new criminal offence can better address culpable behaviour while at the same time better protect freedom of expression. However, it is important to remember that this recommendation is not being made in a legal vacuum: both section 127(1) of the CA 2003 and the MCA 1988 bring within scope many of the behaviours with which we are concerned, and more besides. At the same time, some genuinely harmful behaviours lack criminal sanction. So we are not faced with a choice restricted to a simplistic sliding scale of criminalising more or less (or interfering in freedom of expression more or less). Rather, the existing criminal law is misaligned, and so any new offence should criminalise communications differently, rather than simply more or less. That said, the offence we recommend may well criminalise fewer communications than is currently the case - one of our concerns with the existing offences is the unjustifiably broad scope of criminalisation - but it will also criminalise some harmful behaviour that currently escapes sanction.
2.3 The offence that we recommend is based on a communication’s potential for harm rather than on its content fitting within a proscribed category (such as “grossly offensive” or “indecent” content). We have two particular concerns with the categorybased approach adopted in the existing offences. Our first concern is that the categories lack universally accepted definitions. As we noted in the consultation paper, we do not consider that these terms are capable of unflinching universal definition (what we refer to as “universal standards”), particularly in the case of offensiveness.1 It is difficult for a jury to define offensiveness (or the scale of that offensiveness) without resorting to what they themselves consider to be offensive: in this sense, the terms are susceptible to subjective definition.
2.4 Our second concern is that, without more, neither gross offensiveness nor indecency connote any relevant moral opprobrium: there is nothing inherently wrongful (let alone criminally so) about, say, offensive political satire, or sexual (indecent) communications between consenting adults. Of particular concern is the fact that
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 3.113 to 3.124.
numerous consultees have described genuinely harmful communications sent to them that do not comfortably fit within these categories. Two striking examples of the latter point would include the dissemination of private information or sending illicit surveillance photos of a person to that person.35
2.5 Framing the offence around the communication’s potential for harm is a conceptual shift, and the effect is twofold. First, it ensures that communications that are genuinely harmful do not escape criminal sanction merely because they cannot fit within one of the proscribed categories. Secondly, it ensures that communications that lack the potential for harm are not criminalised merely because they might be described as grossly offensive or indecent etc.
2.6 More fundamentally, however, it moves the focus towards a more context-specific analysis: given those who were likely to see the communication, was harm likely or not likely? Our recommended offence does not require “universal” standards of offensiveness, indecency or obscenity: it does not carry the risk of juries or magistrates importing their own subjective standards by requiring them to define these terms, and instead requires them only to assess the facts and consider whether, on the evidence, harm was likely. We see this as the key strength of the offence.
2.7 The overwhelming majority of consultees supported strongly the move away from the category-based approach. However, a number of consultees - in many cases the same consultees who had supported the move away from the category-based approach - were concerned about the ambiguity that might be introduced by a move towards context-specificity, and the implications this would have for freedom of expression. This is not a concern to be brushed aside lightly - a genuinely vague law would certainly undermine freedom of expression - but we do think the concern is misplaced. First, it assumes a certainty or clarity of definition in the current law that we submit is elusive. Secondly, context-specificity is the solution to the problems associated with the category-based approach and its requirement that juries attempt to define universal standards; there is a sense in which the rejection of universal standards imports - necessitates - context-specificity. One particular way in which this conflict manifests itself is in the calls for the offence to be moderated by appeals to the “reasonable person”, ie it should not be enough that harm was likely based on the likely audience, but that harm must also have been likely to a reasonable person. As we explain below, it is simply not possible to include such a test in the offence without raising the spectre of the universal standards that we have rejected (for the reasonable person test is a form of universal standard).
2.8 In one important respect, however, the offence we recommend differs from the offence that we proposed in the consultation paper. We now recommend that the offence can only be committed where the defendant intended to cause harm. The offence we proposed in the consultation paper could be committed either where the defendant was (i) intending harm, or (ii) aware of the risk of harm. This formulation of the mental element was still a higher threshold for prosecution than under section 127(1) of the Communications Act 2003, which requires only proof that the defendant intended to send the message of the proscribed character; under the current law, the defendant need not have any awareness about any potential result of their actions. Awareness of a risk (an element of recklessness) is also not an unusual formulation in the criminal law. We considered that this was a justifiable interference in freedom of expression because of the need to protect people from harm arising from reckless behaviour.
2.9 Nonetheless, the legal context in which we made the proposal is different from the context in which we now find ourselves. At the time of writing, the Government is considering introducing a Bill that would impose regulatory duties on companies hosting user-generated content (such as YouTube, Facebook and Twitter). 36At the time of publishing the consultation paper, it was not clear what the scope of the regulatory duty would be. If, for example, the duty extended no further than preventing criminal content online, then the scope of the criminal law would have to include all harmful (and sufficiently culpable) behaviour. However, given that the scope of the regulatory regime extends to harmful but legal content, we are now persuaded that an offence based on “awareness of risk” is not necessary to prevent all forms of harmful content and, as a result, might constitute a disproportionate interference in freedom of expression. This is a point to which we return in this chapter.
2.10 One of the implications of narrowing the fault element to one based on intention is that many of the concerns raised by consultees in response to each element of the offence we had proposed, particularly in relation to the scope of the proposed offence, are greatly diminished as they apply to the offence we are now recommending. For example, if the fault element were based on an awareness of a risk of causing serious distress, and had we not included the requirement that the sender lack reasonable excuse, then potentially much political debate would fall within the scope of the offence. Consultees were concerned that “reasonable excuse” was imprecise, but it was the sole element excluding such political discussion from scope. However, by requiring proof that the sender intended to cause harm, genuine political commentary will be excluded from scope by virtue not only of the reasonable excuse element but also by virtue of the fault element. We address these matters in detail in the following.
2.11 In this chapter, we will consider each element of the recommended offence. In the consultation paper, we first asked a question relating to the whole proposed offence, before asking questions about each individual element. This was to help ensure that each element was considered within the context of the overall offence. Many consultees responded to this question instead of - or, in some cases, as well as - the proceeding questions. It therefore makes sense to include the responses to that first question within the relevant sections relating to each element of the offence, rather than having a separate section relating solely to the first question, which would risk significant repetition.
2.12 We should note that we consider there is scope for criminalising communications that are threatening or knowingly false; these aspects of the existing offences ought to persist - albeit in amended form - and we make our recommendations for these offences later in this report in Chapter 3.
2.13 However, we will begin by considering the responses concerning our contention that the existing offences ought to be repealed.
2.14 Our reservations with the current offences are well documented, both in the 2020 consultation paper, and in the 2018 scoping report. In short, we are concerned that the communications criminalised under the current law are not necessarily congruent with harm. Not only is the potential for harm central to the concept of wrongfulness that justifies the criminalisation, we are concerned that the limitations of the existing offences do not include many harmful behaviours.
2.15 In the consultation paper, we noted three broad categories of concern. First, the offences have not fared well in the face of technology. Second, the concept of “gross offensiveness” is imprecise. Third, criminalising indecent and grossly offensive communications is not the same as criminalising genuinely harmful communications (leading both to over- and under-criminalisation).
2.16 As to the first, neither offence applies generally well online, given the manner in which we communicate: the MCA 1988 only criminalises those communications sent to another, so would exclude material posted publicly on, say, Twitter, YouTube, or Facebook; the CA 2003 covers only material sent via a public electronic communications network, so would exclude other forms of communication such as Bluetooth, AirDrop, or a workplace intranet.
2.17 Further, and particularly in the case of gross offensiveness, there is considerable imprecision as to the meaning of the terms. Whilst it might be right to say that, on some specific matters, there might be broad agreement that something is offensive (even grossly so), it is not generally a category that lends itself to easy foresight: someone sending a communication may have little ability to foresee whether a court would find a communication offensive - let alone grossly offensive. People may legitimately disagree on the extent to which a communication was offensive, without either position necessarily being unreasonable or demonstrably wrong. Of course, it is not just the foresight of the potential defendant that matters; vagueness could also lead to inconsistent policing.
2.18 It is also not clear exactly what the justification for criminalising grossly offensive speech is, quite apart from whether we can accurately define it. The words of Sedley LJ in Redmond-Bate v DPP - hackneyed though they may recently have become -are no less true for that fact: “[f]reedom only to speak inoffensively is not worth having.”37 When set against the long history of criminalising grossly offensive communications (at least in the context of public communications networks), those words reveal an uncomfortable juxtaposition in English jurisprudence. For, while the law may go so far as to protect the right to speak offensively, it will criminalise those who speak grossly offensively. Without more, this seems difficult to justify - indeed, it seems contradictory.
2.19 So offensiveness and gross offensiveness do not exist on a sliding scale of wrongfulness but are, instead, two distinct categories in law: one invites protection, the other, criminal sanction. Why should this be the case? It is not clear. If the word “gross” is doing something more than signifying an extreme form of offensiveness -which, to justify the move from protection to criminalisation, it must be - it is not clear what that is. This goes to the heart of our concern with the criminalisation of gross offensiveness: the basis of criminalisation is unclear, and the law is unclear. And this is to say nothing of the imprecision involved in the police, CPS and court determining whether something was offensive in the first place.
2.20 There was substantial support for the proposed repeal and replacement of the communications offences. Consultees strongly supported our criticisms of the existing offences, noting variously the restricted application of the existing offences online as well as the potential both for over- and under-criminalisation.
2.21 The Parliamentary Liaison and Investigatory Team (PLaIT) of the Metropolitan Police Service, in agreeing with the need for reform, noted the current potential for confusion:38
[The] current legislation is too vague and makes the allegations difficult for police officers to keep to a standard nationally, this then causes confusion when one police service in one part of the country takes a case to court and another police service may consider the criminal threshold of abuse/harm isn't met and therefore take no further action... confusion reigns for the police service and for the complainant.
2.22 The National Police Chiefs’ Council similarly noted the potential for inconsistent standards:39
The current model of offences has significant overlaps which leads to inconsistency during investigations and during the provision of charging advice.
The simplification to one primary offence would enable training to be delivered in a focussed manner which will lead to more effective investigation.
2.23 The Criminal Bar Association agreed:40
We agree the existing offences identified under the Communications Act and Malicious Communications Act should be repealed and agree with the concerns about basing criminal liability on what is grossly offensive or indecent, including the ECHR concerns.
2.24 The Magistrates Association agreed with our characterisation of the current communications offences:41
We agree with the detailed arguments put forward in the consultation paper and can see the benefits of creating a new offence to replace section 127(1) of the Communications Act 2003 and section 1 of the Malicious Communications Act 1988. We acknowledge that there is some lack of clarity in respect to the existing situation - especially where certain cases could engage both sections due to overlap, as well as the fact some communications may not be covered by either section. In addition, the lack of consistency between the two sections, especially in relation to technological mode of communication, does currently mean a possible lacuna in respect of some cases.
2.25 Dr Kim Barker & Dr Olga Jurasz commented:42
We agree that there is a pressing need for reform of [section] 127(1) Communications Act 2003 and [section] 1 Malicious Communications Act 1988 so that they are effective, fit for purpose, and adequately capture the nature - and impact - of problematic communications in 2020 and beyond. It is a matter of significant concern that there are a number of gaping holes in the current communications provisions, especially given that this framework does not adequately address online violence, online text-based abuse, nor in particular, online violence against women & girls.
2.26 Recognising the technological limitations of the current law, Zoom also agreed with our rationale for reform:43
We support the modernisation of the Communications Act 2003 and the Malicious Communications Act 1988, in view of the fact that the world has moved online. It is right that legislation keeps up with technological change and modern forms of communication.
2.27 Refuge echoed our concerns with the existing offences:44
In our experience the current criminal offences concerning harmful online communications are not fit for purpose as they do not cover many forms of harmful communications commonly used by perpetrators of domestic abuse and other forms of violence against women and girls. They are also vaguely defined and sometimes poorly understood by police officers and others involved in their enforcement. We support a large number of women whose perpetrators have abused them in ways which we believe fall under the terms of section 127(1) of the Communications Act 2003 and/or section 1 of the Malicious Communications Act 1988, however we see very few investigations and prosecutions for these offences when women report the abuse they’ve experienced to the police. We therefore strongly support reform of this area of law.
2.28 Stonewall agreed that the current communications offences required repeal and replacement:45
Stonewall welcomes the reform and consolidation of the current communications offences, which are piecemeal, inconsistent and complex, into a new communications offence.
2.29 ARTICLE 19 agreed with the need for repeal of the existing offences:46
ARTICLE 19 agrees that section 127 (1) of the Communications Act 2003 and section 1 of the Malicious Communications Act 1988 should be repealed. Indeed, we have been calling for the repeal of these provisions for several years: we intervened in the DPP v Chambers case and subsequently took part in a number of consultations about social media offences and CPS consultations on their guidance on the topic. Specifically, we have long argued for the removal of the terms “grossly offensive”, which in our view is highly subjective and therefore open to broad interpretation with disproportionate consequences for human rights.
2.30 English PEN agreed with the need for reform:47
We support the repeal of the two offences, for the reasons set out in the consultation paper. In particular, that ‘offensive’ communications should not be criminalised... and that communications offences should be ‘technology neutral’. The law should be somewhat ‘future proof’ and what is legal offline should be legal online.
2.31 The Free Speech Union commented on the history of the provisions in the CA 2003 in their response, which supported reform:48
We think this is the correct way forward. We would also point out that [section] 127(1) is a hangover from a previous age, having begun life in the 1930s as a means of dealing with the very specific practice of young men in telephone boxes calling up the (then entirely female) operators for free, and then making obscene remarks to them. It is time it was retired.49
2.32 Carnegie Trust UK echoed our concerns with the existing offences. In doing so, they raised the importance of ensuring that police, prosecutors and judges approach any new offences with consistency:17
Carnegie welcomes the proposal to reformulate the communications offences currently found in [section] 1 Malicious Communications Act [1988] and [section] 127 Communications Act [2003]: it is a parlous state of affairs where provisions of the criminal law manage to be both under- and over- inclusive. In this regard, however, we note also the role of the police, the CPS and the judiciary in recognising the harms caused by speech and flag the need for more training in this area to improve consistency in approach.
2.33 The Media Lawyers’ Association agreed with our analysis:18
The Law Commission expresses concern that the two existing offences under section 1 of the Malicious Communications Act 1988 and section 127 of the Communications Act 2003 are overlapping, ambiguous and so broad as to create a real threat to freedom of expression, particularly as they are adapted to apply to new and unforeseen forms of communications and behaviours. It says that they overcriminalise in some situations, and under-criminalise in others. We agree with these statements.
2.34 Professor Alisdair Gillespie agreed that the current communications offences are in need of reform:19
I agree that [section] 127 and [section] 1 should be repealed and replaced with a single offence. The overlap between the two existing offences is difficult to justify and causes confusion.
There have been concerns (most notably surrounding the Chambers case) that the previous offences are too wide. Thus, the offences are in need of reform...
2.35 The Alan Turing Institute echoed our concerns regarding the relationship between offensiveness and harm:20
.a major problem with the current laws are that they use incorrect terminology, framing what should rightly be called abuse as a question of offence. Whereas ‘offence’ should not be addressed through the law, victims of abuse require protection and the perpetration of abusive acts should, in principle, be punishable.
2.36 What was striking about the consultation responses, aside from the broad support for repeal of the existing law, was the fact that they reflected the full range of our concerns: the vagueness in the terminology, the technical limitations; the limited scope of the offences in the face of online harms, the inappropriate criminalisation of
indecent and grossly offensive communications, and the potential for a prosecution to constitute an interference in Article 10 ECHR.
2.37 Only a very small minority of consultees believed the existing offences should be repealed and not replaced. The majority of consultees agreed with our provisional view that the offences should be repealed and replaced - although there was less consensus at to what should replace them. This is the matter to which we now turn.
2.38 We recommend that section 127(1) of the Communications Act 2003 and the Malicious Communications Act 1988 should be replaced with a new offence. This offence, like the Malicious Communications Act offence, should be triable either-way50 and subject to a comparable maximum sentence. Notably, the offence we recommend includes intention as to result, and in this respect is comparable in form to the Malicious Communications Act offence (and clearly more serious than the section 127(1) Communications Act offence, which requires no intended result and can only be tried summarily51).
2.39 By way of summary, the elements of the recommended offence are as follows:
(1) The defendant sent or posted a communication that was likely to cause harm to a likely audience;
(2) in sending or posting the communication, the defendant intended to cause harm to a likely audience; and
(3) the defendant sent or posted the communication without reasonable excuse.
(4) For the purposes of this offence:
(a) a communication is a letter, article, or electronic communication;
(b) a likely audience is someone who, at the point at which the communication was sent or posted by the defendant, was likely to see, hear, or otherwise encounter it; and
(c) harm is psychological harm, amounting to at least serious distress.
(5) When deciding whether the defendant had a reasonable excuse for sending or posting the communication, the court must have regard to whether the communication was, or was meant as, a contribution to a matter of public interest.
2.40 There are a number of matters that require separate consideration. This section will therefore proceed as follows:
(1) Harm, including the requirement for likely harm;
(2) “Likely audience”;
(3) Intention (the mental element);
(4) “Without reasonable excuse”;
(5) “Sent or posted” a communication; and
(6) Jurisdiction.
2.41 Alongside consideration of the likely audience, the requirement that the communication be likely to cause harm is one of the most significant differences between the recommended offence and the existing offences.
2.42 As we noted above and in the consultation paper, indecency and gross offensiveness are poor proxies for harm. Further, though clothed in the veil of universally understood standards, the proscribed categories are in fact highly subjective, lacking concrete definition and open to inconsistent application. It was therefore our view that, rather than casting about for a proxy for harm that may be more or less accurate on a given set of facts, the law should focus on whether a particular communication actually had the potential for harm. This moves the focus away from attempts to define categories of speech - an exercise that is inescapably flawed - and towards a concrete set of facts. Hence the offence we recommend is a “harm-based” offence.
2.43 There was substantial support for the proposal to shift away from “offensiveness” and similar standards to an approach that sanctions “harm”. Consultees noted the various benefits of such an approach, including the ability to provide greater clarity to investigators, prosecutors and judges (albeit that some had reservations that our offence as formulated could achieve that clarity: we address this point below, and believe that, having now adopted a standard of harm commonly used in the criminal law - serious distress - we address many of these concerns).
2.44 If moving to a harm-based model, there are two further questions that then need to be addressed. First, what do we mean by “harm”? Second, does harm have to be proven (and, if not, what level of potential harm should be proven)?
2.45 We will discuss each of these factors in turn.
2.46 In consultation question 5.1, we asked:52
“Harm” for the purposes of the offence should be defined as emotional or psychological harm, amounting to at least serious emotional distress. Do consultees agree?
2.47 In consultation question 5.2, we asked:53
If consultees agree that “harm” should be defined as emotional or psychological harm, amounting to at least serious emotional distress, should the offence include a list of factors to indicate what is meant by “serious emotional distress”?
2.48 Given that these questions are related, and so necessarily will our analyses be, we will provide a single analysis after we have considered the consultation responses, rather than provide an analysis after each.
2.49 In the consultation paper we drew a distinction between the harm directly flowing from seeing the communication and the more remote (but no less real) harms further along the chain of causation. Our contention was that, with the notable exception of seizures caused by flashing images (a point to which we return in the following chapter), direct harms flowing from seeing harmful communications included psychological harm as the common denominator. Nonetheless, a number of consultees questioned why we had not chosen to include harms beyond the psychological in the elements of the offence. We understand well the wide range of harms that can result from online abuse that extend well beyond the psychological: physical harms and economic harms are two obvious and, sadly, well-evidenced examples. However, first, to the extent that other categories of harm may be direct consequences, we have not seen evidence to challenge our position in the consultation paper that
emotional and psychological harm is the “common denominator” of online abuse. Therefore, a potential new offence would be capable of addressing a wide range of harms flowing from abusive online communications, even if the definition of harm were limited to emotional and psychological harm: the offence would nonetheless cover abusive behaviour that indirectly causes harms...54
2.50 Second, to include all such harms within the scope of the offence would represent a very significant increase in the scope of this field of criminal law. Not only would such an offence overlap entirely with other laws (such as fraud or bribery, in the case of communications causing economic loss) - which alone is reason enough for extreme caution - but it would ostensibly criminalise communications that are not obviously wrongful merely because one indirect result might be that a person, say, lost money. It is important that people know at what point they might be committing a criminal offence, all the more so in the amorphous realm of speech. In order for a person to know when they are likely to commit the offence, the offence must only apply to those harms that are foreseeable as a direct consequence of that person’s actions, ie the effect that flows directly from seeing a communication. Bringing within the scope of the offence those harms further along the chain of causation increases liability significantly, while reducing foreseeability.
2.51 We recommend that harm be defined as at least serious distress. In the consultation paper, we said “serious emotional distress”, where serious meant more than simply non-trivial. In hindsight, the addition of the word “emotional” was unhelpful. Many consultees considered that we were proposing a lower threshold than we actually intended to: it was our intention, by the use of the word serious, to indicate that we were raising the criminal threshold. Currently, the MCA 1988 requires proof that the defendant intended to cause alarm or distress (ie not serious distress), but neither requires proof that any result was likely. In any case, we are not convinced that the word “emotional” added anything that was not already within the scope of “serious distress”.
2.52 As we note in the following, “serious distress” already features in the criminal law (while not differing in substance from the standard we intended in the consultation paper). Adopting a known standard - at least as a starting point - has the dual benefit of clearly aligning the offence to other offences in terms of a hierarchy of seriousness, as well as allowing the courts to use existing experience and understanding. We should be clear that we are not recommending that all existing definitions of “serious distress” be copied immutably to our offence - it is a different offence, and importantly one based on likely harm - though we do consider that these comparator offences will provide valuable grounding.
2.53 The serious distress standard appears in two relevant criminal offences. First, it is one of the definitions of “serious effect” in section 76 of the Serious Crime Act 2015 (controlling or coercive behaviour):
(4) A’s behaviour has a “serious effect” on B if—
(a) • ••
(b) it causes B serious alarm or distress which has a substantial adverse effect on B’s usual day-to-day activities.
2.54 Similarly, in section 4A of the Protection from Harassment Act 1997:
(4) A person (“A”) whose course of conduct—
(a) amounts to stalking, and
(b) either—
(i) ■
(ii) causes B serious alarm or distress which has a substantial adverse effect on B's usual day-to-day activities,
is guilty of an offence...
2.55 Harassment, by comparison, requires a person only to have suffered alarm or distress as a result of A’s course of conduct. By necessary implication, then, our offence adopts a higher threshold of harm (albeit, as we shall see, likely harm) than harassment.
2.56 One of the reasons for adopting a higher threshold of harm is that, unlike harassment (which requires a course of conduct) this offence is complete at the point that a single communication is sent. A course of conduct, being conduct on at least two occasions, arguably requires some degree of deliberation; a one-off communication, even when sent with an intention to harm, may, for example, have been an immediately regretted mistake in a moment of anger. Communications can cause a wide range of emotions, and the scale of communication is almost inconceivable; lest the criminal law take upon itself the task of regulating all communication, the threshold of culpability should be high.
2.57 We nonetheless have been left in no doubt that many communications will still meet that threshold. Many stakeholders have spoken to us about online abuse and bullying that have, at the very least, had “a substantial adverse effect on [their] usual day-today activities...”.
2.58 However, as we note below, consultees expressed strong support for a nonexhaustive list of indicative factors, at least for the purposes of guidance. We consider that this, in conjunction with the above, will go some way to meet the concerns of consultees who were concerned that “serious emotional distress” was too imprecise.
2.59 This proposal to define harm as “serious emotional distress” was contested. Consultee responses varied, with some expressing strong support and others strong opposition. Interestingly, the opposition was not merely from consultees who were concerned that the threshold was too low and would overly impact freedom of expression. Some consultees expressed their concern that the threshold was too high and would not appropriately address victims’ experiences.
2.60 Consultees raised concerns regarding the use of “emotional” distress. A number of legal stakeholders noted that this would be a relatively novel way to categorise harm. In doing so, they contrasted the varying existing tests that require “distress” or “serious distress” without using the term “emotional”.
2.61 Some consultees raised the physical and other harms that can eventuate from online abuse and encouraged their explicit recognition. They acknowledged the argument that emotional or psychological distress may be the “common denominator” but were concerned that this did not sufficiently reflect the nature of the harm that victims suffer. A number of consultees who stressed this also encouraged taking into account other types of harm that may eventuate much later in time from communications, but did not grapple with the difficulties that would arise in proving causation, or in delayed investigation and prosecution, that would eventuate from such proposals.
2.62 The Association of Police and Crime Commissioners agreed with the proposal and noted the existing mechanisms of dealing with other harms directly (eg fraud and harassment):55
We agree with the Commission that for the purposes of this offence, harm should be defined as emotional or psychological harm, given that there are existing offences which cover other harmful online communications, e.g., fraud and harassment.
We believe that the threshold of the psychological and emotional harm should be “serious emotional distress”, and recognise the precedent for this in existing legislation, e.g., section 76 within the Serious Crime Act 2015 which criminalises coercive or controlling behaviour providing that the defendant’s conduct has a “serious” effect on the victim.
2.63 The Magistrates Association agreed with the need to ensure that the scope of the offence is not overly broad:56
We agree that it is important any new offence is not overly broad, and that harms dealt with elsewhere (such as fraud, as explained in the consultation paper) do not need to be covered, and in fact it is beneficial if they are not covered to reduce unnecessary overlap.
2.64 Carnegie UK Trust agreed with the proposal and with our reasoning in the consultation paper. They also emphasised the importance of a “fairly-labelled” offence to tackle online abuse:57
The Carnegie proposal has recognised that the law in England and Wales (whether civil or criminal) does not adequately recognise the impact of speech and the harm that can be caused, especially where the speech could be deemed to fall within the category of political speech. While the Report makes the point that it is important that the proposed offence is not too broad, we emphasise that there is another potentially beneficial side-effect of limiting the harms to emotional and psychological harm - of labelling clearly that words can cause profound hurt.
2.65 Dr Jen Neller noted:58
A remedy to the over-broadness of the proposed offences, additionally or alternatively to specifying that communications are 'abusive', would be to define harm as significant negative impact on daily life. An interference in freedom of expression could be better justified on the basis that it was intended to, or believed to be likely that, the communication would have a significant negative impact on the daily life of a likely audience.
2.66 Both ARTICLE 19 and English PEN disagreed with the proposed standard of harm. English PEN advocated for a requirement of proof of a “recognised medical condition”:59
As stated above (Question 3) we support the introduction of a ‘harm’ element into the law, as an appropriate justification for curbing Article 10 rights. We accept that for communications offences this must be a form of psychological harm to an individual. Incitement to physical violence is dealt with elsewhere in the criminal law, and ‘harm’ to society is dealt with in hate crime laws (the subject of a parallel Law Commission project).
We are concerned that the harm threshold is defined as ‘emotional or psychological harm, amounting to at least serious emotional distress.’ This is lower than the ‘recognised medical condition’ standard found in: civil claims for negligence; the tort established in Wilkinson v Downton [1897] 2 QB 57; and for criminal offences against the person.
We would recommend the same threshold be applied to the proposed new offence, for two reasons.
• First, there is a virtue in adopting a standard that the law already recognised, and for which common law rules are already developed.
• Second, it adds a crucial element of objectivity to the standard. This will prevent trivial or vexatious complaints, just as it does in civil law.
If a ‘likely to’ approach is to be adopted (see Question 4), we are unsure how a Court would be able to assess ‘distress’ without recourse to accepted clinical standards. We fear that in such cases, the ‘serious emotional distress’ standard will become indistinguishable from ‘grossly offensive’ or evidenced by unrepresentative outrage observed in the tabloids or on social media.
2.67 The Crown Prosecution Service urged clarity about the threshold of harm:60
There is some merit in the new harmful communications offence proposed by the Law Commission. However, we would recommend that the term ‘emotional or psychological harm, amounting to at least serious emotional distress’ should be more clearly defined. We therefore suggest that further thought needs to be given to this definition to avoid prolonged arguments at court. Further consideration should also be given to the threshold of serious harm and potential impact on lower-level offences.
2.68 Professor Alisdair Gillespie, in a very full response on the matter, noted that this constituted a significant shift away from the traditional English basis of criminalisation, and would likely result in many “grossly offensive” communications not being criminalised. He called for societal harm to be included in the definition, and further cautioned against setting the threshold too high:61
If the threshold is to be B v New Zealand Police (2017), as set out in paragraph 5.106, then this offence will be used rarely. There, the complainant was unfit for work and upset for a long time. That is a high threshold.
2.69 The Bar Council of England and Wales raised the possibility of including physical harm via shock or alarm:62
We wonder whether the test should also include physical harm caused by shock or alarm. A person may send a communication knowing that the recipient is particularly vulnerable or fragile with intent to cause a harmful physical effect - for example where a victim is known to have a heart condition, or severe asthma. The requirement that the defendant intends harm or knows of the risk of harm to the likely audience will exclude remote or unexpected injury. Otherwise we agree with the minimum threshold of harm suggested.
2.70 There was strong support for a list of indicative factors to indicate what is meant by “serious emotional distress”. While responses varied, generally consultees noted that any list should be distinct from the legislation and offence provision, and that it should be indicative only. The support for this came both from those who supported the proposals and those who opposed them (who viewed the list as a way of mitigating some of the subjectivity with which they took issue).
2.71 Those who opposed the creation of a list of factors did so largely as part of broader opposition to the proposals. They argued that even with a list of factors, the enquiry would remain too subjective.
2.72 The Crown Prosecution Service noted that a non-exhaustive list of factors in statute would be helpful.
2.73 The Parliamentary Liaison and Investigation Team of the Metropolitan Police agreed that there should be a list of factors:63
... it is useful to focus investigators’ minds when they are collecting evidence, not everyone understands why another person may be so offended by a particular post.
2.74 Community Security Trust also supported such a list:64
‘Harm’ is a subjective term and for the legislation to carry public confidence and be applied consistently, its legal definition needs to be set out in clear, comprehensible terms.
2.75 Dr Elizabeth Tiarks & Dr Marion Oswald agreed:65
Yes, we agree that a list of factors would be useful to guide courts and to ensure that the offence is not applied too broadly. It may be useful to consider including both a list of factors that are likely to support a finding of serious emotional distress, as well as factors which would mitigate such a finding. The list of factors in the NZ offence are helpful. It will be important in these factors to reflect recent case-law which has thrown doubt on whether even considerably offensive, irritating, provocative, contentious or abusive communications should be subject to criminalisation (Scottow [2020] EWHC 3421 (Admin); Miller [2020] EWHC 225 (Admin)).
2.76 The Criminal Bar Association argued that a list of factors, while useful, should not be within the terms of the offence itself:66
Yes. A list of factors may be useful, but we consider that may be best left to guidance than the wording of the offence...
Without further definition of the meaning of serious emotional distress, there is a risk that the sense that the Law Commission wish to communicate of this being a ‘big sizeable harm’, as explained in the consultation paper, may be lost. A nonexhaustive list of factors to consider would assist - the Court should have regard to the intensity and duration of the likely distress and how it would likely manifest itself - for example, whether it would have an impact on day to day activities/sleeping-patterns/appetite etc - before concluding that serious emotional distress was likely. This would [aid] the Court to make more objective decisions in cases which may involve inferring the extent of likely distress. A list would also be useful in reducing the risk of unnecessary or inappropriate arrest or intervention by the Police and may enable a more robust and arguably more objective view to be taken regarding a complaint, even in the face of strong pressure by a complainant.
2.77 Stonewall noted that any additional guidance could draw on the New Zealand jurisprudence:67
We support the option presented in 5.114, whereby additional guidance produced could include some of the factors cited in New Zealand jurisprudence (such as intensity, duration and manifestation), to indicate what is meant be ‘serious emotional distress’, while also explicitly stating that this list is indicative and nonexhaustive
2.78 The Magistrates Association noted that a list of factors would likely be of use, including for sentencing:68
We understand the argument of introducing a higher threshold than those set out for offences under the Protection from Harassment Act 1997. It seems a sensible level to set for any new offence. As mentioned in the consultation paper, we have previously pointed out that it is common for courts to judge “seriousness” in different contexts, and we are confident they could do so in respect of this new offence. It may be helpful for factors to be included in any guidance associated with the introduction of the new offence, including sentencing guidelines.
2.79 Professor Tsachi Keren-Paz disagreed:69
I would not support a list of factors as they tend to limit the scope of responsibility (even if they are brought as indicators). Moreover, if it is accepted that emotional harm should suffice - at least as an alternative prong of actual emotional harm -including indicators of ‘serious emotional harm’ will muddle things up.
Having said this, the list of indicators in 5.111 seems reasonable, and in the area of my current research of image-based abuse is typically easily satisfied.
2.80 We recognise that “serious emotional distress” as proposed in the consultation paper risks ambiguity; as noted above, we are of the view that the addition of the word “emotional” only served to confuse matters.
2.81 By aligning the offence more closely with existing law - while recognising that it is a new offence so should not be bound to harassment case law - courts will better understand the nature of the harm to which the inquiry is directed.
2.82 We recognise that distress is often more difficult to prove than physical harm. Of course, this concern has more often been presented to us in the reverse formulation: distress is more difficult to disprove, and this raises the spectre of the vexatious complainant. There are three things to be said about this. First, as far as we are aware, this has not presented an insuperable problem in other criminal laws such as harassment and assault. Secondly, it remains a higher bar than is currently the case: prosecutions under neither of the MCA 1988 nor section 127(1) of the CA 2003 need prove any outcome, real or potential. Thirdly, it is worth bearing in mind that the requirement is for likely harm. We will address this is detail later in the chapter, but for now it suffices to note that the requirement for likely harm imports considerations of foreseeability. The mere fact that someone was harmed (or claims to have been) does not in and of itself mean that harm was likely.
2.83 Nonetheless, we do agree that a non-exhaustive list of factors - in guidance rather than statute - would provide significant help in practice. Like many consultees, we would be reluctant for this to become too restrictive and inflexible: technology changes rapidly and, with it, the precise nature of harms. This is one of the benefits of keeping the factors within guidance rather than statute (and, in any case, we are now recommending a term - serious distress - which carries a more concrete definition in the law of England & Wales than does “serious emotional distress”). However, we do see real benefit in the provision of tangible examples of serious distress, especially at the evidence-gathering stage. The New Zealand factors noted in the consultation paper would seem a fair starting point, with the caveat that they relate to actual rather than likely harm.
2.84 As to the threshold of harm, we recognise that there was considerable divergence on this point. However, it is our view that we have set the threshold appropriately. On the one hand, there are always going to be good reasons to want to lower the bar; harmful behaviour exists below our proposed threshold of criminalisation, of that we have no doubt. However, into that assessment must be weighed practicality: taking WhatsApp alone as an example, nearly three quarters of a million messages are sent every second; 70the criminal law is an important tool in combatting harm, but it does not have the means to address any and all harmful behaviour. Even if it did, one might rightly question whether the criminal law is the right tool for moderating a significant proportion of communication. It is right that the criminal law, with its financial and personal cost, be reserved for the most serious cases.
2.85 That said, we are firmly of the view that defining harm as a “recognised medical condition” is too high a bar. The “recognised medical condition” standard comes from Diminished Responsibility and sets the standard for an abnormality of mind so severe as to render a deliberate act of killing manslaughter rather than murder. 71This would effectively decriminalise a large swathe of communications that are sufficiently harmful as to have a substantial adverse effect on a victim’s day-to-day activities.
2.86 Further, limiting the definition of serious distress to harm which has a substantial adverse effect on a person’s usual day-to-day activities would also be too high a threshold. It is important to note that the statutory offences we mentioned above (controlling and coercive behaviour, and stalking) that use the term “serious alarm or distress” do not define serious alarm or distress as meaning “substantial adverse effect”; they merely use the substantial adverse effect as a magnitude threshold. We are concerned that this threshold, if adopted as the sole definition of serious distress, would impose too high a bar (and note that the comparator offences are much more serious than our recommended offence, having maximum penalties of 10 and 5 years respectively).
2.87 Some consultees noted that the Offences Against the Person Act 1861 (“OAPA 1861”) criminalises at a higher threshold, the suggestion being that our proposed threshold was therefore not worthy of criminal sanction. This is a non sequitur. Merely because some offences are more serious - and the offences in the OAPA 1861 are generally more serious - it does not follow that the Act sets the floor. Indeed, common assault criminalises an offence against the person but at a lower threshold than the OAPA 1861. Further, the OAPA 1861 does not have a monopoly on “recognised standards of harm”. Yet further, to the extent that the contention is that “really serious harm” (which is the standard for grievous bodily harm) is a purely factual inquiry and should therefore be adopted, such a contention fails to recognise that whether serious distress was likely is still a factual inquiry. It does not follow that “serious distress” is rendered capable only of subjective definition by the mere fact of being a different threshold. In any event, many criminal offences criminalise at a much lower threshold of harm: assault, battery, criminal damage, theft, fraud, causing harassment, alarm or distress,72 not to mention the existing communications offences. Of course, it is true that expression warrants particular consideration, though it is then not immediately clear that the OAPA 1861 is the best comparator.
2.88 Nonetheless, it is worth bearing in mind that the scope of the offence is constrained by other factors that the prosecution has to prove, such as intention and lack of reasonable excuse. Concerns about breadth do not all have to die on the hill of harm.
2.89 Consultation question 4 asked:73
We provisionally propose that the offence should require that the communication was likely to cause harm. It should not require proof of actual harm. Do consultees agree?
2.90 Though the majority of consultees supported this proposal, it was not without controversy.
2.91 There are three important points to note. First, requiring proof of actual harm for a relatively low-level offence is not particularly common: again, neither assault nor battery require proof of actual harm. Secondly, as we have mentioned before, a person should be able to foresee whether their actions will be criminal. If the offence were complete at the point of actual harm rather than likely harm, the only point at which the defendant would know an offence had been committed is when a person is harmed. An actual harm element makes sense when the victim is present at the time an offence is committed: a punch could cause no harm at all, or Actual Bodily Harm, or Grievous Bodily Harm, or even death. The problem with a communications offence is the gap between when the offence is committed and when the harm is caused.
2.92 Thirdly, part of the rationale was not to expand the scope of the offence - as many believed - but instead to constrain the scope of the offence. The risk of the vexatious complainant under our model is lower than is currently the case: a victim might wish to testify that they found something to be grossly offensive, and that might be enough to fortify a court’s finding that, indeed, a communication was grossly offensive. Further, the risk of the vexatious complainant is not precluded by an actual harm element. The mere fact that a person claims to have been harmed does not in and of itself mean that harm was likely. We say that culpability rests on the fact that harm was likely and not the mere fact that, through some chain of causation, somebody was actually (or claims to have been) harmed. Our submission is that the “likely harm” criterion is more restrictive and more normatively significant than a requirement that the offence cause actual harm, however unlikely.
2.93 Finally, some consultees suggested that it wasn’t clear what level of probability accorded with “likely”. We potentially over-estimated the comprehensibility of this word when we published the consultation paper.
2.94 Professor Alisdair Gillespie agreed that an approach requiring actual harm would be problematic:74
Restricting the offence to that where actual harm took place would be problematic where there is the ‘resilient victim’, i.e. while D intended to cause harm, V actually just shrugged off the attack. As noted in paragraph 5.84(1), it would also mean that the police do not need to find proof of actual harm. That could be problematic where, for example, a public tweet was sent.
2.95 Professor Jacob Rowbottom also agreed with the proposal, and in particular noted that this standard “avoids difficult questions of causation and media effects.”75
2.96 The Criminal Law Solicitors’ Association agreed: “If the communication is such that it is obvious that there was a likelihood of causing harm then it is sufficient in [our view] that there should be no proof of actual harm. It is the intent to cause harm which is relevant.”76
2.97 The Parliamentary Liaison and Investigation Team of the Metropolitan Police also noted the importance of ensuring an offence is not avoided merely due to a “resilient victim”: “If the intent to cause harm is there^the offence should be complete. Just because a person decides not to be hurt by a communication doesn't make the sender innocent”.77
2.98 The Crown Prosecution Service agreed and observed that the position, rather than being novel, is similar under the existing communications offences:
We agree. This is a helpful proposal, and we note that the current communications offences under [section] 127 CA 2003 and [section] 1 MCA 1988 do not require proof of actual harm either. This is in order to include scenarios where no actual harm occurred or could be shown. For example, this may apply when posts do not target at a specific recipient and in cases where no victim comes forward. As the Law Commission outlines, harm to the information subject is addressed in the civil law of defamation.
2.99 Community Security Trust agreed, explaining the importance of being able to address communications that an intended recipient deliberately avoids reading:78
If, for example, a person deliberately sends an antisemitic communication to a Jewish recipient, intending to cause them fear and distress, they should not be protected from prosecution if that Jewish recipient anticipates their intention and does not read the offending communication.
2.100 The Association of Police and Crime Commissioners agreed with the approach:79
The APCC agrees with this proposed approach. As the Commission rightly assesses, proof of harm may be especially difficult to obtain in the case of public posts, where it is difficult to establish who has seen the communication.
Equally, as the commissioners of victims services locally, we believe that victims should not have to go through the potentially re-traumatising process of producing evidence that they were harmed, although evidence of actual harm could count towards proving that a communication was likely to cause harm.
2.101 Stonewall agreed and provided evidence of the types of abuse LGBT+ communities suffer online:80
Stonewall supports the removal of the requirement to prove actual harm in law, in line with the Commission’s rationale that ‘Proof of harm may be especially difficult to obtain in the case of public posts, where it is difficult to establish who has seen the communication. Further, given the culture of dismissiveness and normalisation around harmful online communications, victims may be especially unlikely to come forward’ (5.84 - 85).
2.102 Refuge noted that this aspect of the proposals is important in ensuring any offence is workable in practice:81
Refuge strongly agrees with this element of the proposed offence. Requiring proof of actual harm would create an unnecessarily high evidential barrier to prosecution and is likely to result in limited use of the offence. As the Law Commission acknowledges in the consultation, proving harm was caused by a particular communication can be very challenging, particularly in the context of domestic abuse and other forms of VAWG where harmful communications are highly likely to be only one of many ways a perpetrator is abusing a survivor as a pattern of coercion and control.
2.103 Nevertheless, consultees also pointed out that the “likely harm” test was not without its difficulties when it came to proof. Alisdair Gillespie commented:82
...That said, it will be challenging to prove the likelihood of harm. In para 5.90 you indicate that an actual victim is harmed would help, but if you have harm suffered by an actual victim then you don’t need to prove likely harm. Factors may help (discussed later), but it is still going to be difficult to prove. The Public Order Act
1986 is different because it is easier to prove that someone is likely to be harassed or alarmed. Showing the likelihood of serious distress appears quite challenging.
2.104 The Magistrates Association discussed proof in the context of the Public Order Act 1986:83
We agree with the proposal, for the reasons stated in the consultation paper. We note the query as to how a prosecution would prove a communication was likely to cause harm. As stated in the consultation paper, evidence of a victim being harmed may be determinative in providing proof, but also note that a similar test is currently used in relation to offences under section 5 of the Public Order Act 1986.
2.105 The Law Society of England and Wales questioned how the offence would operate in practice:84
When deciding whether the communication was likely to cause harm to a likely audience, the court must have regard to the context in which the communication was sent or posted, including the characteristics of a likely audience.
This therefore tries to put the communication into context, practically, for the judge or magistrates and asks the court to examine the context of the sending of the communication and the characteristics of the likely audience; but it is still an objective test and there is a risk of inconsistency in the decision-making.
We consider that it will be necessary for there to be very clear guidance and references to examples so that the judiciary understand the context of messages, to understand what was intended and what harm was really caused.
2.106 We take seriously all of the concerns regarding proof. Consultees raised convincing arguments that both the actual harm test and the likely harm test would present difficulties of proof.
2.107 However, we are satisfied that the appropriate test is one of likely harm. We reiterate the point above that it is important the defendant be able to predict at the point of sending whether they are about to commit a criminal offence. Further, the mere fact that someone was harmed does not imply that harm was likely (which has bearing on the culpability of the defendant). The jury or magistrate will have to determine as a matter of fact that, at the point of sending, harm was likely. If a person has an extreme and entirely unforeseeable reaction, the element of likely harm will not be satisfied. This, combined with the problems associated with requiring proof of actual harm in certain commonplace circumstances (such as public posts in public fora), as well as the difficulties in proving causation (identified by Jacob Rowbottom), fortify our view that the benefits of such an approach outweigh the difficulties associated with proving likely harm.
2.108 We further think that those difficulties can be mitigated by clear and appropriate CPS guidance. It is perhaps somewhat difficult to consider in the abstract, but part of the rationale for the harm-based offence is context-specificity and particular scrutiny of the facts.
2.109 For the reasons stated above, we are not convinced that the test of likely harm raises problems for freedom of expression or increases the chances of vexatious complainants.
2.110 As to the question of what “likely” means, our view is that it is not enough that there is a mere risk or possibility of harm. We consider that this would invite too great a degree of interference in expression by the criminal law. 85Conversely, requiring proof that harm was “more likely than not” would be an inappropriately high bar; this sets the threshold of culpability very high and would present real problems of proof. Instead, our view is that there needs to be a “real or substantial risk of harm.” This was the test adopted by the House of Lords in Re H for the purposes of defining the Children Act 1989 standard of “likely to suffer significant harm”.86
2.111 Finally, a number of consultees have suggested variations on a theme of the “reasonable person” as a way to mitigate what they regard as excessive subjectivity in the harm test. This argument was raised in response to a number of questions, but it is as well to raise it here. There is a particular difficulty with this argument.
2.112 The reasonable person test requires a notion of what harm a reasonable person might have been caused when faced with a particular communication. Who is the reasonable person? The reasonable person is not simply the “majority”; it is a more abstract notion than that (the majority may well be unreasonable). One of the chief problems with the reasonable person test in this context is that it is not clear which attributes the reasonable person must share with the victim. If they share only the fact of being human, and maybe being adult, and maybe speaking the same language, then it is not clear what the test is doing other than replicating the “universal standards” inherent in the categories in the existing offences (the very thing that consultees accepted was unascertainable and an undesirable basis for the criminal law). What if, in the alternative, the reasonable person shares some more of the characteristics of the victim: their age, gender, disability, socio-economic background? All are characteristics that might have legitimate bearing on how someone might respond to a communication. In that case, it is not clear what the reasonable test achieves that the test of likely harm does not: if the reasonable person is, in effect, the victim, we are instead just asking whether harm was foreseeable, or likely, given the nature of the communication and the victim.
2.113 This works both ways, of course. If a person knows the likely audience of their communication shares their dark sense of humour, there is no reason why that person should be found guilty because a more “reasonable” person would likely have been harmed. Alternatively, if the sender knows the likely audience is unusually susceptible to a particular type of harm and exploits that to their serious disadvantage, the sender should not be able to argue their innocence by appeals to the reasonable person.
2.114 It is also important to stress that there are other constraints within the offence: the sender will have to have intended harm, and also be proven to lack a reasonable excuse. If a person intends harm, and has no reasonable excuse for that, it is not clear why that person should not have to “take their victim as they find them” (which is the basis of quantifying damage in civil law). If, however, a recipient were harmed in a way that a sender did not foresee, then it is very unlikely that the sender intended that harm. We therefore do not think that a reasonable person test would add anything to the existing offence.
2.115 Consultation question 3 asked:87
We provisionally propose that the offence should require that the communication was likely to cause harm to someone likely to see, hear, or otherwise encounter it. Do consultees agree?
2.116 We should begin by noting one important mitigating factor that has arisen as a result of our removal of the “awareness of a risk” limb of the offence. Whilst a notable majority of consultees supported this proposal, a number disagreed. One of the reasons for this was because, evidently, the net is cast somewhat wider than under the MCA 1988 (which requires a communication be sent to a person); a person might be harmed by encountering a communication (or by seeking it out) even though the defendant did not intend that person to see it. Concerns were particularly acute in the realm of online political debate. However, although that person may still be in the “likely audience” for the purposes of this element, to be guilty the defendant must have intended harm to the likely audience. Necessarily, then, the defendant must have foreseen, at least in broad terms, who that audience would be and have intended them harm. For the avoidance of doubt, it is not enough that, say, an old email was discovered that someone now considers might likely cause harm; the defendant must have intended it be harmful when they sent it.
2.117 As far as the “likely audience” test is concerned, it is necessary to make two preliminary points. As formulated in the consultation paper and in this report, the offence is complete at the point the communication is sent. Therefore, the assessment of who was likely to see the communication is to be assessed “as at” the point of sending. This is not the same as asking who the immediate recipient was - it is broader than that - but the mere fact that the original message was shared, or the old Tweet sought out some years later, would not imply that that was likely.
2.118 A related point, and it is one we also make in respect of likely harm, is that “likely” here means a “real or substantial risk”. 88It does not simply mean “possibility” or “risk”.
2.119 Taken together, these points curb significantly the scope of liability. To be guilty, the sender would have to intend harm to those who were at a real or substantial risk of “seeing, hearing, or otherwise encountering” the communication.
2.120 Alisdair Gillespie agreed with the proposal, arguing:89
It should not be restricted to those the communication is sent, as that would cause difficulties, particularly in the online environment. It should not be a defence to use an ‘open’ form of communication and say, “ah, but I only intended X to read it”.
...Of course, as you note, this does mean that private conversations are (now) exempt. Currently, if two people engage in an obscene conversation online then, at the very least, it constitutes an offence under [section] 127. That will not be the case now. The argument in favour of such an approach is that two people in a pub having the same conversation would arguably not be prosecuted. The courts have previously justified the distinction where it involves the abuse of a publicly-funded communication network (DPP v Collins (2006)). That argument is arguably no longer tenable, and so the proposal to de facto decriminalise conversations save where the recipient suffers, or is likely to suffer, harm is perhaps more justifiable.
I agree with the suggestion in paragraph 5.80 that who is likely to receive the messages must be an objective test. The nature of the internet means that a subjective test would be problematic. Where a person is sending a message that is capable of harming another, it is not unreasonable for him to understand the potential audience of his [message].
2.121 The Free Speech Union agreed with the “harm-based” approach. In their view, the approach adopted in New Zealand does not have sufficient safeguards for expression:90
We certainly agree that this requirement must form the outer boundary of any offence. The New Zealand provision concerning information distressing to the subject of it seems to us to go much too far in creating a general crime of making distressing comments about other people.
2.122 Refuge noted the strength of the proposals in being able to recognise harms that occur outside of the “communicator/recipient” relationship:91
Refuge welcomes that the proposal that the offence should apply when the communication was likely to cause harm to someone likely to see, hear or encounter it. It is a real strength of the model that it is not restricted to communications directly aimed at the victim-survivor. If the offence were to be limited in this way, it would fail to include many of the ways in which perpetrators of domestic abuse and other forms of VAWG perpetrate abuse through communicating publicly on a particular forum or platform, or communicating to a friend, family member or colleague.
2.123 The Bar Council of England and Wales agreed with the proposal, but stressed the importance of what they characterise as the “qualifying element” of intent:92
We agree. This is dependent on the qualifying second element, namely that the defendant “intended to harm, or was aware of a risk of harming, a likely audience”. Without that intention or knowledge it would be too wide and uncertain.
2.124 Mr Justice Martin Spencer, responding in his capacity as chair of the Judicial Security Committee, echoed these sentiments: “Liable to cause harm would be better, this keeps it in line with other criminal offences. Also ‘intention’ should remain, as it is the core principle.”93
2.125 The Association of Police and Crime Commissioners agreed with the proposal and of our analysis of the impact of online abuse from the consultation paper:94
The APCC agrees with this proposed approach, and the Commission’s assessment that the harm of online communications can extend beyond those towards whom the communication is directed.
2.126 Though a number of consultees were content that, in the majority of instances where information is shared about a person online, that person would be in the scope of the offence, some consultees called specifically for “information subjects” to be within the scope of the audience. For example, Suzy Lamplugh Trust argued:95
...the legislation should explicitly include victims (information subjects)96 who have been affected by the communication, but may have not likely seen, heard or otherwise encountered it.
2.127 By contrast, Dr Elizabeth Tiarks and Dr Marion Oswald discussed alternative means of protecting information subjects:97
The removal of the words “or otherwise encounter it” would narrow the offence to a more reasonable scope. This does not leave indirect victims without redress, as there are alternative methods of tackling indirect harms to “information subjects” through civil law, and specifically the tort of misuse of private information and data protection remedies, which may well be more appropriate to example 1 on page 113 of the consultation document.
2.128 Consultees have fortified our provisional view that restricting the scope of the offence to “recipients” would, in the context of modern communications, be too narrow. As we note above, it is our view that the requirement to prove that the defendant intended to harm the likely audience constrains the scope of the offence to justifiable limits (a point echoed by the Bar Council in their response). Equally, it would seem unusual to allow a person to escape liability for onward sharing that was entirely foreseeable to them (they may, for example, have even wished or directed it).
2.129 One issue that arose in the consultation responses was whether “likely audience” should be defined to include “information subjects”. What this means is, if a communication is about someone, but they are not otherwise likely to see, hear or encounter it, whether that person should still be defined as falling within the “likely audience”.
2.130 We do see the force in that argument but, as we noted at para 5.79 in our consultation paper, this would constitute a dramatic expansion of the scope of criminal law. It would almost certainly, for example, re-criminalise defamation. There is undoubtedly harm that flows from the sharing of personal information, but the gravamen of the offence is different from a communications offence (which is directed at the harm to those who witness a communication). These two may overlap - and consultees have agreed that, in some instances such as “doxing”, the communications offence may be applicable - but alternative means of criminalising the disclosure of private information, or criminal defamation, are distinct and discrete legal questions that do not lie within the scope of this report.
2.131 Nonetheless, by contextualising the harm to the likely audience, it does allow the assessment of likely harm to be set against the particular characteristics of that audience. We consider this a vital element in the proposed offence, and is the matter to which we now turn.
2.132 Consultation question 6 asked:98
We provisionally propose that the offence should specify that, when considering whether the communication was likely to cause harm, the court must have regard to the context in which the communication was sent or posted, including the characteristics of a likely audience. Do consultees agree?
2.133 Ultimately, this question reflected our view that harm is inherently context dependent -that what harms one person may be entirely harmless to another. Indecent messages may, in one context, be genuinely harmful (cyberflashing is a good example of this) whereas the same messages passing privately and consensually between two people is clearly not criminally wrongful conduct (quite apart from questions of intention).
2.134 There was strong support for this proposal. By and large, consultees supported our analysis in the consultation paper that set out the importance of having an offence that can take the context of communications into account. Consultees emphasised the importance of considering communications in their proper context rather than in isolation. Many consultees also noted that being able to recognise and address context is central to ensuring that any offence is effective in addressing the harms that victims experience.
2.135 Some consultees’ responses seemed to refer to “protected characteristics” as found in hate crime legislation. This confusion may have arisen from the use of the phrase “characteristics of a likely audience.” We did not mean “characteristics” to have so prescriptive a definition.
2.136 Regarding freedom of expression, there was some divergence among consultees. Some saw this proposal as central to protecting freedom of expression, others saw it as exacerbating the problems they had already raised.
2.137 Many consultees stressed the importance of having an offence that can properly consider the context of harmful communications and the impact they have on victims. Stakeholder groups provided detailed evidence about the impact of online abuse on the people they work with and emphasised how context-driven proposals would better address this than the existing offences.
2.138 A number of consultees noted that this proposal will better allow the intersectional experiences of those who are subject to online abuse to be recognised.
2.139 Demos emphasised the benefits of a context-specific approach in the context of online communication:99
We agree that the offence should make these specifications and be attentive to the context of the communication, including the characteristics of a likely audience. The harms of online communication can be complex and highly specific to a context, meaning that those unfamiliar with the context may overlook how a harm is being perpetrated.
2.140 Fix the Glitch set out their view that any offence must look both at the content and context of communications:100
We agree that the court should examine the context in which the communication was sent or posted, the court should examine the context in which the communication was sent and the characteristics of a likely audience. Online abuse and violence is not solely dependent on the content of the communication made; in some cases, harmful communications can be deemed so because of the context in which they were made.
2.141 Refuge gave examples of communications whose harmful nature is not objectively evident on the face of the communication, but highly context-dependent:101
Refuge strongly agrees that the context of the communication must be considered when establishing whether the communication was likely to cause harm. This is particularly crucial in the context of domestic abuse and other forms of violence against women and girls, where a particular communication might not seem obviously aggressive or threatening to an observer, but when looked at in light of the pattern of abuse carried out, it is highly distressing and harmful.
Examples of communications made to women supported in our services include perpetrators tracking down survivors after they have fled their abuse to a secret location and sending them a picture of their front door or street sign on their road. Whilst a picture of a front door isn’t harmful in most contexts, in the case of domestic abuse it is deliberately sent to cause intense fear and distress. Such communication also has significant implications for women, for example if they will often need to move again to different area in order to be safe.
2.142 Women’s Aid also noted the context of violence against woman girls (VAWG):102
Experiencing harmful online communications in the context of domestic abuse is also likely to have a significant impact on victims. As the Sentencing Council Guidance on Domestic Abuse makes clear, “the domestic context of the offending behaviour makes the offending more serious because it represents a violation of the trust and security that normally exists between people in an intimate or family relationship. Additionally, there may be a continuing threat to the victim’s safety, and in the worst cases a threat to their life or the lives of others around them. ”103 Similarly, victims who are targeted by harmful communications in the context of sexual violence, harassment and other forms of VAWG will experience particularly severe impacts. For these reasons, the context of domestic abuse, sexual violence and other forms of VAWG must be specified as particularly harmful contexts for the court to consider, and data collected on the offence should be routinely disaggregated by offender and audience characteristics, and their relationship.
2.143 The Crown Prosecution Service agreed and referenced the approach in New Zealand to contextual factors:104
We agree. We note New Zealand's Harmful Digital Communications Act 2015 includes a non-exhaustive list of contextual factors including the extremity of the language used, the age and characteristics of the victim, anonymity, repetition and the extent of circulation. Although we recognise that there is a risk that including a list of contextual factors might not capture all relevant factors in every scenario, we would welcome the inclusion of such a non-exhaustive list in the proposed statute.
2.144 The Association of Police and Crime Commissioners agreed with both the proposal and our analysis that prompted it in the consultation paper:105
We agree that the court must have regard to the context in which the communication was sent or posted, including the characteristics of a likely audience, and agree with the Commission’s recognition that “context is often key to understanding how and why a communication is harmful”.
2.145 The Magistrates Association agreed with our characterisation of this proposal as reflecting the “established approach” of courts. They noted further: 106
... that other jurisdictions list relevant factors in similar legislation, but agree with the Law Commission’s judgement that this is not necessary in this case, as courts should have discretion to respond appropriately by taking all relevant factors into account.
2.146 English PEN agreed that this proposal is central to protecting freedom of expression:107
Context is crucial and must be considered by the court. The argument in the consultation paper (at para 5.119) is that a simple requirement on the Courts, to have regard to the context in which the communication was sent, should allow for sensible decisions. Coupled with the special protections for freedom of expression set out at section 12 of the Human Rights Act 1998, c.42, it is not unreasonable to assume that the Courts will deliver adequate protection for human rights, when presented with a case at trial.
However, as discussed above we are wary about the overdue sensitivity of the audience being a critical factor in a prosecution.
2.147 The Free Speech Union stressed the importance of a context-specific approach in safeguarding freedom of expression:108
We entirely agree. In our respectful view, one of the objectionable features of the case of Markus Meechan, mentioned in connection with Question 5 above, was the apparent holding of the Sheriff, upheld in the Sheriff Appeal Court, that, in deciding whether exhibiting on YouTube a video of a dog doing a Nazi salute was “grossly offensive”, the overall context - clearly that of an intentionally bad-taste joke - was irrelevant.
2.148 Some consultees noted their concerns about the extent to which this proposal would allow any offence to be “weaponised” by a person who claims to be the victim of online abuse. Many of these consultees expressed their concerns in reference to advocating for women’s organisations and single-sex spaces.
2.149 The LGB Alliance disagreed. While they noted that the proposal may seem reasonable, they argued that the practical reality of online mass communication makes it unwieldy:109
While this may seem reasonable, it creates major difficulties on such platforms as Twitter which are open to all. The likely audience is everyone in the world. Given the ease of mass communication, it is a simple matter for a lobby group, such as those asserting the importance of gender identity theory, to organise as part of any audience - making them a ‘likely’ part of that audience. If they so desire, they can assert the right of transwomen to be potential partners for lesbians and claim that those who disagree are harming them. This is not a fictitious example: it occurs on a regular basis.
2.150 Sex Matters disagreed. They noted that a context-driven approach would “give too much discretion to police and CPS”. 110They provided a detailed response discussing at length the various impacts of existing offences have had on gender-critical academics and others who seek to advocate in favour of single-sex spaces.
2.151 We take seriously the concerns of those who believe the context-specificity of the offence could be used in a way that runs contrary to the legitimate exercise of Article 10 rights. However, in demonstrating the profoundly divergent emotional responses to communications, the arguments serve primarily to demonstrate that any liability based on universal standards will be unworkable.
2.152 However, contentious political discussion is one of the reasons why we think the “without reasonable excuse” element is so fundamental (to which we will turn shortly). Notably, the court must consider whether a communication was or was intended as a contribution to a matter of public interest.
2.153 The other important factor that will provide protection to freedom of expression is the mental element of the offence: where no harm was intended (such as, for example, might be the case in political argument), the offence cannot be made out. Therefore, while we accept arguments - such as those made by the LGB Alliance - that people could “organise themselves” to form part of the likely audience, there are other constraints within the offence that will prevent the criminalisation of ordinary (even if highly contentious) political discourse.
2.154 Some of the examples provided by consultees demonstrating the context-specificity of harm were striking - such as the example of the photograph of the front door provided by Refuge - and echo some of the examples of which stakeholders have spoken throughout this project. These have fortified our view that the model of offence we have proposed is appropriate.
2.155 As noted at the beginning of this Chapter, the offence we recommend should require proof that the defendant sent a communication intending to cause harm to the likely audience. In the context of criminal law, where it is a result that must be intended, intention means that a person did an act “in order to bring about” a particular result.111 It is to be distinguished from recklessness (which requires merely an awareness of a risk of that result). This was not what we proposed in the consultation paper: the offence that we provisionally proposed could be made out by proving either an intention to cause harm or awareness of the risk of causing harm.
2.156 We considered that this dual form of mental element was necessary because we had seen - and have continued to see - many examples of harmful behaviour where it is not clear that the defendant intends harm. Importantly, not all of this is wrongful behaviour - hence the requirement that the prosecution also prove a lack of reasonable excuse. However, some of the behaviour clearly is wrongful, such as the people whom we have heard send flashing images to people suffering from epilepsy “for a laugh”.
2.157 Nonetheless, the context of the criminal law is important, and that context has shifted since we published our consultation paper. The Government is likely to introduce an Online Safety Bill112 that would impose a duty of care over organisations hosting usergenerated content. That duty of care will obviously not cover communications that take place offline (or even off the respective platforms), but the nature of communication is such that it will cover a very significant proportion of harmful communications. It is also our view that platforms are in a far better position to prevent harmful content than the criminal law; criminal justice is an expensive tool with limited resources to deploy.
2.158 However, and this is crucial, our recommendation that the offence should require intention only and not awareness of risk in the alternative is based on the existence of a regulatory regime that covers some form of legal but harmful content. If the duty of care is limited to the scope of the criminal law (ie platforms only have a duty to prevent criminal offences online), or simply doesn’t exist, then the criminal law would have to step into the void.
2.159 Consultation question 8 asked:113
We provisionally propose that the mental element of the offence should include subjective awareness of a risk of harm, as well as intention to cause harm. Do consultees agree?
2.160 One thing worth noting is that the responses occasionally describe the two forms of the offence as the “intentional” form and the “reckless” form. For the avoidance of doubt, recklessness requires proof that the defendant was aware of the risk (in this case, risk of a result, that result being harm) and that, in the circumstances as known to the defendant, it was unreasonable to take the risk.114 This is a common law definition. We separated these elements in the offence - rather than including the word “reckless” - because of the very particular role played by reasonableness in the context of this offence and how careful we needed to be in defining it. We therefore considered it important not to leave the scope of the offence to be affected by any change in the definition of recklessness.
2.161 Victim and law-enforcement stakeholders generally expressed strong support for the proposal, arguing that the more flexible an offence the better it can ensure that all culpable behaviour is addressed. Consultees also by and large agreed with our analysis that a “reckless” form of the offence would address behaviour not currently addressed by the existing communications offences.
2.162 A range of concerns were raised, primarily concerning freedom of expression and potential over-criminalisation. Stakeholder groups with particular interests in freedom of expression noted their concern that including awareness of a risk as a fault element would have too great an impact on communications and would likely “chill” expression to a substantial degree. In particular, that in combination with the “likely harm” and “likely audience” criteria, and the threshold of “serious emotional distress”, the lesser fault element would criminalise large swathes of speech that are not currently criminal.
2.163 Further, legal stakeholders (for example the Law Society of England and Wales) noted that the potential breadth of an offence with awareness of a risk as a fault element may criminalise unintentional behaviour and have a disproportionate impact on young people. Their concerns were heightened given the nature of online communications, where the impact of a communication is more difficult to gauge, and the effort required to send or post it is minimal. The potential consequences of a criminal conviction might significantly outweigh the culpability.
2.164 In contrast, Professor Tsachi Keren-Paz outlined his view that awareness of the risk of harm would be a preferable threshold for the mental element:115
Absolutely. A major weakness of [section] 33 of CJCA 2015 [“revenge porn”] is its insistence on intention to cause distress. The analysis in the consultation paper is robust.
2.165 Young Epilepsy outlined some of their own experience of receiving malicious communications and their preference for a mental element requiring awareness of a risk of harm: “The malicious communications we have received imply that the perpetrators have (at the very least) a subjective awareness of the risks associated with sharing flashing images with the epilepsy community.”116
2.166 Alisdair Gillespie discussed the importance of addressing behaviour beyond merely intentional acts:117
Including recklessness is important if it is going to be a harm-based offence. Proving that a person knowingly sent a communication intending it to cause serious emotional distress could be difficult. Recklessness is common with harm-based offences. There will still be the requirement that a person knowingly or intentionally sends a communication, and so the recklessness only applies to the harm element.
2.167 Refuge commented a broader threshold would make the offence more workable in practice:118
Yes, Refuge agrees that any new offence should include subjective awareness of risk of harm as well as intention to cause harm. Such an approach would lead to a more workable offence that would capture circumstances where evidence of intention is challenging, but where it was clear that a perpetrator was aware their communication could cause harm but were reckless or disinterested in this risk.
2.168 The Association of Police and Crime Commissioners outlined their view that culpable “reckless” behaviour should be addressed:119
We agree with this approach, and the Commission’s assessment that including “subjective awareness of a risk of harm” within the offence would be consistent with previous recognition within law of culpability when clear intention to harm is not present, e.g., in the legal concept of “recklessness”.
We also agree that including a subjective awareness of a risk of harm could help to prosecute instances of cyber-flashing, where the perpetrator’s primary intention may not be to cause harm. We believe that focusing too narrowly on the perpetrator’s intention, will draw focus away from the risk of harm to a likely audience, and therefore dilute the proposals’ overall “harm-based” approach.
2.169 The Crown Prosecution Service noted that while the existing offences require intention, the recognition of “recklessly” culpable behaviour is common in the criminal law:120
We agree. While [section] 1 MCA 1988 and [section] 127 CA 2003 require intent, many other existing criminal offences already incorporate degrees of culpability from intent to recklessness to correspond with a scale of sentencing options. This approach has worked well in other offences such as the Protection of Harassment Act 1997.
The Law Commission considers that in some circumstances, abusive communications should be criminalised even if the motivation is not to cause harm. The anecdotal experiences of CPS prosecutors tend to support this insight. This approach may help to avoid the technical issues that arise for the prosecution where the defendant’s motivation may include other factors such as ‘having a laugh’.
2.170 The End Violence Against Women Coalition argued that this proposal could ensure a wider range of perpetrators are held to account:121
We support the Commission’s proposal that the mental element includes not only an intention to cause harm, but an awareness of the risk of harm (para 5.148). This will help to ensure that a wider range of perpetrators are held responsible for their abusive acts. It will also make prosecutions more likely as demonstrating proof of a direct intention to cause harm is particularly challenging in the online environment where there is often little additional material or context other than the abusive communication.
2.171 The Magistrates Association accepted our analysis in the consultation paper and supported the proposed lower fault element:122
We appreciate that including the second test of the mental element that the defendant should be aware of a risk of harm does widen the scope of the offence. However we note the Law Commission’s intention not to criminalise harmful but not culpable behaviour, and support their argument that introducing this test will ensure those who were aware of the likely harm of their communication, even where it cannot be proven they intended harm, are potentially culpable. It is also important that those who are fully aware of the likely harm due to a communication, but where intended harm is not the primary driver for the behaviour, are held to account.
2.172 The Criminal Bar Association agreed that a scope beyond full intention was appropriate, but argued that it would be more appropriate to express this as “recklessness” rather than as a mere “risk”:123
Yes, we agree that some subjective element is necessary and we understand the offence as currently drafted requires D to be subjectively aware of a risk of harm as defined in the section i.e. serious emotional distress, not just a risk of any harm. We consider for the reasons set out above (see paragraphs 29-38) that the subjective awareness of risk of harm necessary for an offence to be committed should be expressed as ‘recklessness’ rather than mere ‘risk'.
2.173 However, a number of consultees were concerned about the potential for over-criminalisation. For example, the Law Society of England and Wales raised their concerns about the potential for over-criminalisation if the proposed “lower” fault element was adopted:
It has been suggested by practitioners that immature teenagers, in particular, would not see the risk of harm. There is a considerable danger to young people who, isolated in their rooms and among their group of friends, may say things that are ill-thought out or are, in fact, intended to shock. Many of these young people will never have been in trouble in their lives and their futures might be blighted by a criminal conviction for something they will never have realised was an offence.
There is a danger of imposing a current moral norm on someone and for that it is suggested there would need to be an objective test - the ordinary reasonable person on the Clapham Omnibus. The question would become whether or not a reasonable man would the see the risk of harm.
2.174 Dr Elizabeth Tiarks & Dr Marion Oswald noted that the proposal may be overly broad and referenced their recommendation to require a “significant risk” of harm in any lower fault element.124
2.175 The Justices’ Legal Advisers and Court Officers’ Service (formerly the Justices’ Clerks’ Society) argued that “awareness of risk” of harm was a disproportionately low fault element for a relatively serious criminal offence:125
We can see why you would wish to create something wider than intention, for the reasons you give. However, we think “awareness of a risk” of harm is too low a threshold for criminal intention for such an offence with such high consequences.
For comparison, and appreciating the different purposes behind the legislation, it certainly wouldn’t be considered appropriate to add it into section 4A Public Order Act 1986 for example.
2.176 Certain consultees also had concerns about the implications for freedom of expression flowing from a relatively low mental threshold. Kingsley Napley noted their concerns: “We believe that subjective awareness may go too far. To preserve freedom of speech, we recommend that the offence is limited to the intended harm.”126
2.177 The Free Speech Union disagreed and set out their preferred approach:127
... Most people engaging in robust speech are likely to be aware that what they say may cause severe distress to at least some people. In the absence of any intent to cause distress, we are not convinced that that feature alone should be sufficient potentially to criminalise their speech.
2.178 ARTICLE 19 set out their preference for requiring “full” intention rather than awareness of a risk of harm:128
...it would have a significant chilling effect on public conversations on contentious topics since the risk of harm and therefore awareness of that risk would be inherent to the topic. To give an example, an ordinary user sharing disturbing footage about a terrorist attack for journalistic purposes would almost inevitably upset [victims’] families. The truth can often be seriously distressing.
2.179 English PEN accepted that some form of recklessness could be appropriately captured, but that it must not disproportionately impact freedom of expression:129
We agree that there should be a mental element to the proposed new law. When a message has been targeted or addressed to an individual, then it would be appropriate to include a ‘recklessness’ limb as an alternative to direct intent.
However, when a message is broadcast to the world on a blog or social media, an ‘awareness of risk’ or ‘recklessness’ limb would be inappropriate. Due to the high likelihood that controversial or offensive messages can cause ‘serious emotional harm’ to someone on the internet, a ‘risk of harm’ limb would become trivial to prove, and effectively turn the offence into one of strict liability.
2.180 Under Article 10 of the European Convention on Human Rights (ECHR), interferences in freedom of expression should be necessary. We are bound to consider the ECHR implications of laws that we recommend. We acknowledge that an offence based on awareness of a risk constitutes a greater (arguably, a quite considerably greater) interference in freedom of expression than does the intentional offence. In one important sense, however, we have built Article 10 protection into the offence, in requiring proof that the defendant lacked reasonable excuse (and it is this test which imports consideration of political debate). Nevertheless, we have to consider whether there is a lesser interference that could achieve the legitimate aim, especially considering the weighty burden of justification for imposing criminal sanction. To this end, we cannot ignore the regulatory regime that the Government is planning to introduce in respect of online harms.
2.181 We have sympathy with those consultees who argued that there was clearly scope for an offence based on awareness of risk: in one sense, we entirely agree. Where we might differ is whether the most effective or appropriate response is a criminal one. Given the significant work being done by the Government to improve safety in the context of online communications, and given the scale of online communications as a medium of communication, we are of the view that a criminal law response to deal with merely reckless communications, rather than intentionally harmful ones, is not necessary (and, by extension, not proportionate).
2.182 We equally have sympathy with those consultees who questioned the implications for children and young adults, and for freedom of expression, of having an offence based on awareness of a risk of harm. (We also address these matters in the introduction of this report, at paras 1.50 to 1.57).
2.183 The chief benefit of requiring proof of intention is that it helps to constrain the scope of the offence to the genuinely culpable, relieving the “without reasonable excuse” element of much of its burden. For example, a doctor delivering bad news to a patient - where serious distress may well be likely - is absolutely not in the scope of the offence; quite apart from the fact that they obviously have a reasonable excuse, they have also not acted in order to bring about any harm (indeed, they almost certainly regret the harm that they are causing). A similar argument could be made in response to ARTICLE 19’s example concerning images of a terrorist attack shared for journalistic purposes.
2.184 We recognise that it will be more difficult to prove intention than it would have been to prove that the defendant was merely aware of a risk. However, we do not think that it will be an impossible burden: proof of intention is a very common requirement in criminal offences. Often, the content and context of a communication may well be enough for the jury to infer that the defendant must have intended harm (even if the defendant claims otherwise). Consider, for example, a person sending images of corpses in a concentration camp to an elderly Jewish person: the defendant in that case would struggle to marshal any credible defence to the contention that they intended harm.
2.185 The jury can also consider the defendant’s foresight of the natural and probable consequence of a communication as evidence that it was intended. However, and importantly, it is absolutely not the case that intention can be established “so long as” the result was a natural and probable consequence of the defendant’s conduct, or even so long as the defendant foresaw the result as a natural and probable consequence. Evidence of the natural and probable consequence (or the defendant’s foresight thereof) may be evidence of intention, but this is not the same as saying that such evidence alone will inevitably suffice to establish intention. As we note below, for the natural and probable consequence to suffice as evidence of intention, the likelihood of that consequence (ie the degree of probability) would have to be overwhelming.
Direct and indirect/oblique intention
2.186 In order to understand how evidence of the natural and probable consequence of conduct informs the question of intention, it is first necessary to distinguish between direct intention and indirect (or oblique) intention. These are the two “processes” for establishing intention.
2.187 A person will normally be taken to have intended a result if they acted in order to bring that result about. 130This is direct intention, to be contrasted with indirect intention. For the latter, a person may be taken to have intended to bring about a result if that
person thought that the result was a virtually certain consequence of their action.131 However, even if the result were a virtually certain consequence, the jury does not have to find that it was intended; they may nonetheless decide that a result clearly was not intended (as would be the case in the example of the doctor above).
2.188 Either definition of intention can suffice for a finding that a defendant had the necessary intent to be guilty of an offence. However, it is rare that a direction to the jury concerning foresight and indirect intention would need to be given: it is only necessary in those situations where the evidence would suggest that the defendant’s motive was not to bring about the relevant result.132
Natural and probable consequences
2.189 Evidence of the defendant’s foresight of the consequences may form part of the evidence that they acted in order to bring about the result (ie evidence of direct intention), but it will only suffice to establish intention (in this case, indirect intention) if the defendant foresaw the consequence to a very high degree of probability.
2.190 It is never enough to point to the natural and probable consequences of conduct and, on that basis alone, prove intention. This principle was clearly established by the House of Lords in Moloney. 133Juries, when considering whether there is evidence of intention, may draw inferences from the natural and probable consequences of the defendant’s actions, but this is not the same as saying that intention can be established so long as the result was the natural and probable consequence of the defendant’s conduct. The question for the jury is whether the defendant intended the outcome (or whatever matter to which the intention need be directed) based on an assessment of the evidence. In establishing direct intention, the defendant’s foresight of consequences might be evidence that they intended them, and the natural and probable consequences of their conduct might be evidence of that foresight, but “foresight and foreseeability are not the same thing as intention”; they are matters of evidence and not legal presumption.134
2.191 In his speech in Hancock, Lord Scarman remarked:
.. .the greater the probability of a consequence the more likely it is that the consequence was foreseen and that if that consequence was foreseen the greater the probability is that that consequence was also intended.135
2.192 However, for such foresight to suffice in establishing intention, the probability must be high indeed. On this point, Lord Scarman cited with approval the part of Lord Bridge’s speech that considered probability:
..the probability of the consequence taken to have been foreseen must be little short of overwhelming before it will suffice to establish the necessary intent.136
2.193 This was confirmed subsequently in Nedrick 137and Woollin. 138In Nedrick, concerning the intention element of murder (the defendant must intend death or serious injury), Lord Lane CJ held that:
... the jury should be directed that they are not entitled to infer the necessary intention, unless they feel sure that death or serious bodily harm was a virtual certainty (barring some unforeseen intervention) as a result of the defendant’s actions and that the defendant appreciated that such was the case.139
2.194 This approach was confirmed by Lord Steyn in Woollin, who was at pains to draw the fundamental distinction between intention and recklessness. In Woollin, the trial judge had directed the jury that they could infer intention so long as they were sure that the defendant foresaw a “substantial risk” of serious injury. Lord Steyn held that this was a material misdirection: substantial risk was clearly a wider formulation than virtual certainty, and thus unacceptably blurred the line between intention and recklessness.140
2.195 It is therefore not correct to suppose that a finding that a consequence was merely probable would lead inexorably to a finding that it was intended. Further, we must emphasise that even a virtually certain consequence does not lead inexorably to a finding of intention: the jury may nonetheless find that there was no intention. So a doctor delivering bad news to a patient, or a manager making an employee redundant, or a professor telling a student that they had failed a module, may know that their communications will lead to serious distress, but that does not mean that a jury would find that they intended that serious distress (and, in any event, they are all outside the scope of the offence because none of them lacks a reasonable excuse).
2.196 We also considered whether framing the fault element in terms of the defendant’s purpose was preferable. We concluded that it was not.
2.197 “Purpose” is more restrictive than intention: it speaks of a person’s motive, their reason for action, the outcome they desire. Proof that a person foresaw a result as probable, even highly probable, will not be enough to sustain a finding that they desired the outcome. The terrorist might place a bomb in a car park and provide fair warning to help ensure that the car park is evacuated: their purpose is not to kill (albeit that, when then bomb detonates and kills another, we would have no problem saying that they intended to kill or cause serious injury).
2.198 However, purpose is not a particularly common way of defining the fault element in crimes. Sometimes it is entirely appropriate, and there are certainly examples.141 Two would include: section 1 of the Official Secrets Act 1911 (which makes it an offence for a person to do certain acts for a “purpose prejudicial to the safety or interests of the State...”); and section 67A of the Sexual Offences Act 2003 (which prohibits “upskirting” for the purpose of obtaining sexual gratification or causing alarm, distress or humiliation). Nonetheless, defining the scope of a person’s purpose is not always an exact science. For example, in Chandler v DPP, 142several persons organised a “ban-the-bomb” demonstration at a Royal Air Force base and were convicted of conspiring to commit a breach of section 1 of the Official Secrets Act 1911. The House of Lords upheld their conviction on the basis that their relevant - or immediate -purpose was not to get rid of nuclear weapons (albeit that that was their object), but rather to obstruct aircraft. Whilst this inexactitude will likely present few problems in respect of defining a purpose for sexual offences - the range of ulterior purposes beyond the sexual or abusive for taking non-consensual private, sexual images of another may be somewhat limited - a purpose does become increasingly difficult to establish where people’s reasons for acting may be as many and varied as they are in the field of communications.
2.199 Our concern with purpose in the context of the harm-based offence is not only that it may be evidentially difficult to establish, but also that it is likely to be too restrictive. It is not difficult to imagine a most abhorrent and genuinely distressing communication where a defendant might claim that their motivation wasn’t to cause serious distress, but where serious distress was so overwhelmingly likely that the culpability was still high. If, for example, a person sent flyers to nursing homes suggesting that their residents be forcibly euthanised - and did so with a view to saving the state money (or at least a stated view to that effect) - it seems that their culpability is not so far removed from that of the person who wrote the same messages desiring serious distress as the outcome. Purpose, then, would fail to capture some genuinely harmful and - importantly - culpable communications.
2.200 It is for a similar reason that we should be slow to adopt for our offence the wording from section 47(7)(b) of the Serious Crime Act 2007, which (as with section 44(2) of the same Act) was intended to preclude a finding of indirect intention:143
D is not to be taken to have intended that an act would be done in particular circumstances or with particular consequences merely because its being done in those circumstances or with those consequences was a foreseeable consequence of his act of encouragement or assistance.
2.201 On its face, of course, section 47(7)(b) says nothing that is not already trite law. 144As we note above, evidence of foresight might strongly favour a finding of intention (the
“irresistible inference”145), but the mere fact that consequences were foreseen does not lead to a legal presumption of intention. The wording of section 47(7) - “taken to” - serves only to preclude a presumption of intention based merely on foresight; it does not preclude the use of foresight as evidence of intention. Even indirect intention requires proof of more than merely foreseeable consequences: the consequences must be foreseen and virtually certain (and, of course, foreseen as such).
2.202 However, even if this section is interpreted in such a way as to preclude a finding of indirect intention, we do not recommend adopting it for the purposes of the harmbased offence. To exclude a finding of intention based on virtually certain outcomes, where evidence of the defendant’s motive may otherwise be in short supply, would make this offence very difficult to prosecute, and would fail to capture culpable behaviour.
2.203 We are therefore satisfied that intention is the most appropriate formulation for the fault element of the offence. It is sufficiently narrow in scope that it addresses concerns about an overly broad offence, without presenting an insurmountable evidential hurdle for the prosecution.
2.204 This was one of the most widely misunderstood parts of our consultation paper, though also one of the most important. Many consultees thought that we had proposed a defence of “reasonable excuse”, ie the burden would fall on the defendant to prove that they had a reasonable excuse, lest they be found guilty of the offence. This was not what we proposed. Instead, our proposal was that the prosecution should be required to prove as part of the offence that the defendant lacked a reasonable excuse - and that needed to be proven beyond reasonable doubt.
2.205 It was for this reason primarily that we considered the offence complied well with Article 10 ECHR. Unless the court was sure that the defendant lacked a reasonable excuse, considering also whether it was the communication was or was intended as a contribution to the public interest, the defendant would not be found guilty.
2.206 It is important also to recall that we are recommending a different and narrower fault element for this offence than we had proposed in the consultation paper. We originally proposed that the offence could be made out if the defendant were aware of the risk of harm or intended harm. We now recommend that the offence can only be made out where the defendant intends harm. The range of communications for which there might be a reasonable excuse is far smaller where harm is intended than where harm is foreseen as a risk. For this reason, many of the concerns raised by consultees (below) in relation to the reasonable excuse element are greatly diminished or simply do not arise.
2.207 Alisdair Gillespie’s response articulated the tension between a safeguard forming part of the conduct element of an offence (which is important) and terming the safeguard an “excuse”: the latter appears to frame the issue in a way that puts the burden of proof on the communicator, even though our analysis in the consultation paper and reasoning for including the safeguard seek to ensure that the onus is on the prosecution. However, the term “reasonable excuse” does have some advantages over simply asking whether the communication was (or was not) “reasonable”. Whilst reasonable excuse allows some scope for assessing the content of the message as well as the reason for its being sent, the assessment of whether a communication was reasonable necessitates an assessment of the content. It requires assessment of whether the communication was itself reasonable - ie does the content meet a standard of reasonableness - quite apart from whether the excuse for sending the communication was reasonable.
2.208 Take the example of a satirical piece of writing on a matter of intense political interest. Under our recommendations, this is very unlikely to fall within the scope of the offence (not least because of the requirement that the sender intend harm): the prosecution would have to prove that the defendant lacked a reasonable excuse; the defendant would argue (successfully, we submit), that political satire is a perfectly reasonable excuse for sending a communication that may cause a person harm. 146However, if the question were instead whether the piece of satire was reasonable, it is not at all clear by what metric we could answer this question. Of course, we could consider such matters as who was likely to see it or the likely harm, but these are already elements of the offence, so it becomes an exercise in circular reasoning. Many communications are simply not amenable to a “reasonableness standard”. We therefore remain of the view that “lack of reasonable excuse” is the better formulation, despite the risk that it be viewed incorrectly as a defence.
2.209 We were also aware that “reasonable excuse” outside of a specific context might be vague. It is hard to impose prescriptive factors that will apply in all situations; the circumstances in which the issue might arise are infinite. Indeed, “reasonable excuse” does not have any precise definition in law. In R v AY,147 the Court of Appeal held that “[t]he concept of ‘reasonable excuse’ is par excellence a concept for decision by the jury on the individual facts of each case.”148
2.210 However, one category seemed to call for particular consideration when determining whether a communication was reasonable, and that was whether it was or was intended as a contribution to a matter of public interest. If an exercise of freedom of expression bears on a matter of public interest, the ECtHR will be less ready to find that an interference - especially a serious interference like a criminal conviction - will be compatible with Article 10.149
2.211 Consultees expressed very strong support for this proposal. A minority were concerned that such a consideration could be used by defendants as an attempt to “justify” otherwise clearly harmful content, though we should note that the public interest does not override the operation of the reasonable excuse element; it is simply a matter that must be considered. A person may be communicating on a matter of public interest, but that does not lead inexorably to a finding that the person had a reasonable excuse. Given the nature of many of the responses on this point, which did not appear to acknowledge this feature of the proposal, it is clear that we could have expressed this more explicitly in the consultation paper.
2.212 Other consultees noted the imprecise nature of the public interest. However, assessment of the public interest in the field of communications is not unknown to the law. For example, under the section 4(1) of the Defamation Act 2013, it is a defence to an action for defamation for the defendant to show that the statement complained of was, or formed part of, a statement on a matter of public interest. Courts have nonetheless been reluctant to constrain the concept in precise ways; as noted by Warby LJ in Sivier v Riley, “it is common ground that [the notion of public interest] is necessarily a broad concept.” 150In any case, we reiterate that the public interest consideration does not mandate a finding one way or the other as to the reasonable excuse element; this provision is simply intended to ensure that public interest matters are at the forefront of the court’s mind. This reflects the importance that the European Court of Human Rights attaches to speech concerning matters of public interest.
2.213 We will first note the responses concerning the reasonable excuse element, before then considering the responses relating to the public interest part of that element. We will present a combined analysis at the end of this section.
2.214 Consultation question 11 asked:151
We provisionally propose that the offence should include a requirement that the communication was sent or posted without reasonable excuse, applying both where the mental element is intention to cause harm and where the mental element is awareness of a risk of harm. Do consultees agree?
2.215 Consultees expressed strong support for this provisional proposal. Even those who opposed the provisionally-proposed offences noted that, should they be taken forward, they should include a requirement they be sent without reasonable excuse as a “crucial safeguard”.
2.216 A substantial number of legal stakeholders agreed with our analysis in the consultation paper and set out various ways in which the concept of “reasonable excuse” was well-known to the law. Further, many noted the greater protection afforded by ensuring that “reasonable excuse” is not a defence, but rather something to be disproved to the criminal standard.
2.217 The Justices’ Legal Advisers and Court Officers’ Service (formerly the Justices’ Clerks’ Society) noted their preference for a number of separate offences arranged in a hierarchy, similar to the scheme in the OAPA 1861. Chara Bakalis made a similar observation in her general response. Each notes the difficulties that arise with our provisionally-proposed offences in relation to what constitutes a sufficient “harm” and what may amount to a “reasonable excuse”. They argue that the provisional proposals try to act as “catch-all” offences, and that it may be more appropriate to adopt instead narrower, more targeted offences to address the varying types of harm.
2.218 Despite supporting the safeguard itself, consultees expressed concern with the potential vagueness of the concept “reasonable excuse”. A number suggested that a non-exhaustive list of factors could be used to assist in giving more content to the concept. However, as noted by the legal stakeholders, “reasonable excuse” is a relatively familiar concept to the law and to courts. To that end, it may well be useful to have some form of guidance as to what may constitute a “reasonable excuse” that is not in legislation: possibly in specific police material or CPS guidance. Consultees commended the guidance under the New Zealand model in this context.
2.219 On the topic of “reasonable excuse” being known to the law, the APCC’s response referencing the use of the concept in coronavirus restrictions and regulations152 is worth noting. The concept has been used extensively as a way to ensure there were adequate safeguards in place when the coronavirus restrictions were applied. However, there have been numerous examples of inconsistencies in approach by police. 153To that end, the potential formulation of police guidance or similar resources to assist in consistent application may be useful.
2.220 The Justices’ Legal Advisers and Court Officers’ Service (formerly the Justices’ Clerks’ Society) agreed and discussed a number of the examples used in the consultation paper:154
We agree that reasonable excuse is a necessary requirement in the proposed offence. Your examples are interesting.
Take the example of the ending of a relationship. People end relationships in sometimes very traumatic ways. If someone sends an electronic message saying “it’s over,” clearly that is not criminal in the proposed offence. They have a reasonable excuse (i.e. autonomy to live their life how they see fit).
However, consider if they said: “it’s over, I hate you. I’ve hated you for a long time. You were useless at a, b and c, and I only stayed with you because of x, y and z” (i.e. they deliberately said things to hurt the other person). Perhaps accompanied by obscenities or reference to very personal things. If those words were spoken in the home to the partner, none of that would likely be criminal.
Yet in the proposed offence, it is sailing close to the wind; the hurtful comments are arguably not said “without reasonable excuse” (there being no good reason to say them other than to cause upset), they could cause emotional distress, but surely we do not wish to criminalise such comments when put in messages, yet not when spoken?
This is one of the issues with the harm based approach we feel. The New Zealand list of guiding considerations is therefore useful.
2.221 Fair Cop agreed:155
We support this, as presumably it will go to protect Article 10 rights in that contribution to a genuine and necessary debate will presumably be seen as a 'reasonable excuse'.
2.222 Alisdair Gillespie agreed with both the importance of a “reasonable excuse” safeguard and also its inclusion as part of the offence itself as opposed to a defence. He discussed some of the possible issues that may arise in the context of communications that are not inherently “wrongful”:156
Stating that the communication must have been sent without reasonable excuse could be seen as an essential safeguard. If adopted, I also agree that it should constitute part of the actus reus of the offence and not a defence. This is important not only because it clarifies that the prosecution need to prove the absence of a reasonable excuse, but also because of the logic of offences.
A defence is only required where the prosecution have proven the actus reus and mens rea. In other words, a person is prima facie guilty unless they have a defence. That would be wrong where a person had a legitimate reason for sending the communication. By including it within the offence, it is clear that there is no presumption that they have acted wrongly, and need to prove that they did not act inappropriately.
However, the terminology is arguably problematic. ‘Reasonable excuse’ still carries the suggestion that a person should not have said what they did. It is an excuse, a way of showing that the wrong is not culpable. Yet will there be a wrong?
I know that my student has had the maximum number of condoned fails permissible. If she fails my 5,000 word essay then she cannot be awarded a degree.
Unfortunately, the essay is rubbish. I send the feedback and the low mark to her. This causes severe emotional distress because she realises that she has no degree after four years of work.
Why do I need an excuse to send this feedback? Surely, it is a legitimate thing for me to do? Does it form a matter of public interest (see next response)?
2.223 The Association of Police and Crime Commissioners agreed that this aspect of the offence will ensure that both “legitimate” communications and freedom of expression more broadly are protected:157
We agree that the offence should include a requirement that the communication was sent or posted without reasonable excuse. This will help to ensure that legitimate communications which could risk causing emotional distress are not criminalised (e.g., sending a formal communication regarding a change in legal status, such as refusal of housing, etc.).
We also believe that including the reasonable excuse element will help to ensure that freedom of expression is protected, and recognise the previous examples in law where the reasonable excuse concept appears, the most recent example being the Health Protection (Coronavirus, Restrictions) (England) Regulations 2020.
2.224 English PEN emphasised that the “reasonable excuse” aspect of the proposals is an essential safeguard in ensuring an appropriate scope. They argued that examples of what a “reasonable excuse” might be should be set out either in the offence itself or in a schedule:158
The requirement is essential for limiting the scope of the law. If sub-section (3) were absent, the proposed law would amount to a disproportional interference with Article 10.
We recommend that examples of ‘reasonable’ excuse be set out within the offence, or within a schedule.
The Law Commission’s parallel consultation on hate crime analyses the explicit protections for freedom of expression afforded by the Public Order Act [1986] c.64, at [section] 29J and [section] 29JA. The proposed new communications offence should include an analogous provision.
2.225 The Free Speech Union noted the importance of this proposal in ensuring appropriate protection for freedom of expression. They specifically supported including the justification as an element of the offence itself:
We enthusiastically agree. The criminal law has no business whatever sanctioning people who say things which they have a reasonable excuse to say.
We would add that in our view the absence of reasonable excuse should be a substantive element in any offence, rather than its presence a specific defence. Hence it is our view that any burden on the defendant should be merely evidential: provided the defendant raises some evidence of reasonable excuse, it should then be up to the prosecution to prove its absence beyond a reasonable doubt.
2.226 The Magistrates Association agreed and noted the relatively common and straightforward use of “reasonable excuse” in the criminal law:159
We agree that any new offence should include a requirement that the communication was sent or posted without reasonable excuse. Especially if the second mental element test is that the defendant was aware of a risk of harm, rather than a likelihood of harm. Even if the second test is that the defendants was aware of a likelihood of harm, it is still important that behaviour that is likely to cause harm is not criminalised where there is good reason for someone carrying out that behaviour. The examples given in the consultation paper show that there are occasions where someone will send or post a communication that they are aware is likely to cause harm, but where there should be no criminal liability (for example when breaking up with someone or informing them of bad news). We agree that the phrase “without reasonable excuse” is straightforward, and therefore arguably preferable to a recklessness element.
2.227 The Bar Council of England and Wales agreed and noted that in the majority of matters this will be a straightforward enquiry:160
We agree. The circumstances should make this obvious in most cases. Where a potentially reasonable excuse exists (or is raised in evidence by the defence) it must be for the prosecution to disprove it. The Law Commission point out in §5.163 the type of communication which foreseeably causes distress but is a necessary part of social life, such as a doctor’s assistant notifying a person that their child has been diagnosed with a serious illness. By making a reasonable excuse part of the offence to be excluded by the prosecution, the proposed definition avoids causing unnecessary invasion into personal lives (and so avoids offending against Art.8 ECHR).
2.228 Kingsley Napley argued that should the offence be limited by requiring “full” intention, the “reasonable excuse” safeguard may not be needed:161
If the offence is limited to an intention to cause harm then this may not be necessary as an individual would not be guilty unless they intended harm to be caused. If they did intend to cause harm to a ‘likely’ (or in our preference, ‘intended’) audience then it is difficult to comprehend what the reasonable excuse would be...
2.229 Mr Justice Martin Spencer, writing in his capacity as chair of the Judicial Security Committee, argued that the preferable course would be to require a defendant to prove “reasonable excuse” as a defence:162
In criminal law generally, the law has two approaches to reasonable excuse, one where it is raised as an issue by defence - the prosecution has to prove it was used e.g the example you give.
The alternative is that the defence has the burden of proving on balance of probabilities that he had reasonable excuse and relies on that defence, e.g having an offensive weapon in public place. We prefer the second line where online communication is liable to cause serious harm to likely audience, he should have the burden of proving he has the reasonable excuse.
2.230 Refuge raised their concerns about the potential exploitation or manipulation of any “reasonable excuse” safeguard by perpetrators of abuse:134
While we understand the intention behind the inclusion of a without reasonable excuse requirement, we are concerned that, unless tightly defined, this requirement could be manipulated by perpetrators of abuse to evade criminal responsibility for their actions.
Perpetrators of domestic abuse and other forms of VAWG will often try and justify and excuse their abuse as ‘showing they care’, for example the specialist tech abuse team at Refuge often hear perpetrators try and justify the constant surveillance, monitoring and control of women by saying that they love and care about them so need to know what they are doing and where they are at any time.
2.231 Jacob Rowbottom cautioned that “once the threshold of the offence has been met, the expression is more likely to be seen as ‘low value’ and the reasonable excuse clause will be applied sparingly.”135 Further, Rowbottom suggested that many cases would be challenging on their facts:
For example, a parent with strong religious beliefs may attempt to persuade his or her daughter not to have an abortion by sending a message that includes images of an aborted foetus. If the court concludes that the communication is likely to cause severe emotional distress to the recipient, would such an attempt at persuasion be reasonable or not?
2.232 Consultation question 12 asked:136
We provisionally propose that the offence should specify that, when considering whether the communication was sent or posted without reasonable excuse, the court must have regard to whether the communication was or was meant as a contribution to a matter of public interest. Do consultees agree?
2.233 As we note above, there was very strong support from consultees for the proposal that any reasonable excuse element should include consideration of the public interest.
2.234 The Crown Prosecution Service agreed with the proposal, emphasising the importance of ensuring protection for freedom of expression:137
This aims to strengthen protection for freedom of expression under Art. 10 ECHR (in addition to reasonable excuse). The Law Commission considers that in matters of public interest, many of which may be sensitive, there should not be a criminal sanction for communications even if they may cause serious emotional distress.
This fits with the proposed new offence seeking to exclude news media, broadcast
136 Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 12.
media and cinema. Broadcast media were also excluded from s.127 (4) CA 2003 and criminal libel was abolished in 2010.
We note there is a range of case law on matters of public interest, particularly with regard to defamation and libel cases against newspapers. It is also set out as a defence in the Defamation Act 2013 s.4.
2.235 Mermaids also agreed, noting the range of important other factors that would weigh in any assessment of the reasonable excuse:163
We agree that the proposed new offence should specify that, when considering whether the communication was sent or posted without reasonable excuse, the court must have regard to whether the communication was or was meant as a contribution to a matter of public interest, as it is vitally important to protect freedom of expression in a democratic society... We would support the case-by-case nature of this proposed provision in that the extent of the ‘likely harm’ caused should be balanced against the right of freedom of expression, freedom from discrimination and right to a private life.
2.236 Hacked Off agreed that “genuine efforts” to engage on matters of public interest should not be criminalised:164
We broadly agree with the position as set out. Genuine efforts, such as those described in the consultation paper, to engage on matters of public interest should not be criminalised. The test should, however, be an objective one. It should not be a defence for a person to have a subjective but unreasonable belief that they are contributing to a debate of public interest.
2.237 Demos supported the proposal, but noted in an extensive response that there were complexities in establishing the public interest:165
We agree that the offence should make these specifications, and we believe the court should be given extensive guidance in this regard to reflect the complexity of considering public interest here. An area of our recent research has highlighted, for example, how harms can occur even in a context of apparent public interest that is addressed in the Law Commission’s consultation paper.
[I]t is important for the Law Commission to be cognisant that efforts to identify people as engaging in such communications could also run the risk of discriminating, for example, on a class or socioeconomic basis, if certain ways of speaking are deemed unsuitable for ‘civil’ public debate by people in positions of power; a risk that exists too when relying on algorithmic detection of harmful communications. It is vital that we do not create a two-tier system, punishing those whose speech is not aligned with an algorithmic or computational definition of ‘civility’ rather than those whose speech indeed causes significant harm.
[S]ome guidance for the court would be essential in this domain as the nature of communication online continues to present new challenges to judgments of when something is or is not a genuine contribution to a matter of public interest.
2.238 Similar sentiments regarding the approach to political debate were echoed by Fair Cop, who argued that the bar must not be set too high:166
[A]s Knowles J commented in Miller v College of Policing, Mr Miller's tweets may have been in the main 'unsophisticated' but they were nevertheless part of a genuine and necessary political debate. This defence should not be restricted only to those who are articulate or refined in their language
2.239 Jen Neller, albeit framing our proposal as a defence, noted the importance of excluding from scope abusive communications that happened to concern a matter of public interest:167
...such a provision would need to be drafted carefully to ensure that 'contribution to a matter of public interest' would not be interpreted as definitive of a reasonable excuse, but rather as merely one possible consideration in that determination.
2.240 English PEN and the Free Speech Union both argued that the public interest test should form part of an explicit and separate part of the offence, quite apart from the reasonable excuse element. English PEN noted that this was especially important in the context of journalistic activity, noting that there should be a “clear lane through which such communications may pass uninhindered.”168
2.241 The Law Society of England and Wales suggested specifically including a reference to the ECHR in the provisions:169
We agree [with the proposal], but would suggest adding words specifically mentioning the need for the court to have regard to the ECHR in considering the question of reasonable excuse. We are mindful of the Anti-Social Behaviour Crime and Policing Act 2014 pursuant to which, when a local authority is making a Public Spaces Prohibition Order. s 72 thereof governs the decision-making process and specifically directs the decision-maker to consider the rights contained in the ECHR.
2.242 A small minority of consultees disagreed with the proposal. ARTICLE 19, for example, argued that, though well-intentioned, the “public interest” safeguard was insufficient (though it is important to stress that we were not proposing limiting the definition of reasonable excuse to public interest matters):170
ARTICLE 19 understands that the proposal to include a requirement for the courts to have regard to the question whether the communication was meant as a contribution to a matter of public interest is aimed at protecting freedom of expression. This proposal goes some way towards achieving that aim. In our view, however, it is insufficient...
Freedom of speech should not be understood as the freedom to say what judges or prosecutors believe to be in the category of ‘public interest’ or things that have a ‘reasonable excuse’. People often express ideas in ways which are thoughtless, callous or indifferent to people’s feelings. The Law Commission should be very wary of proposing a new offence that would criminalise everyday expression almost by default.
2.243 It is worth clarifying how we see this element of the offence working. A communication will not be within scope of the offence unless it is a communication for which there was no reasonable excuse. Even then, this offence is not concerned with any communications for which there is no reasonable excuse: the communication would have to be likely to cause harm to a likely audience, and the sender must have intended to cause harm. So this element provides a safety valve to ensure that, even when someone has sent a likely harmful communication intending to cause harm, there is no interference by the criminal law in reasonable communications or legitimate freedom of expression. Indeed, because the burden of proof is on the prosecution, the criminal law will only intervene if a court is sure that the sender lacked a reasonable excuse.
2.244 As we noted above, it is not possible to be prescriptive when defining a reasonable excuse in this context. This is easier to do if there is a blanket ban on a particular activity where the law permits specified exemptions. An example of this might be “do not leave your house unless one of the following reasons applies”. The range of possible reasons for speech, however, is so broad that any attempt to define those reasons according to which might be reasonable would inevitably be too restrictive. As the Bar Council rightly noted, the circumstances should make it obvious in most cases whether or not there was a reasonable excuse; it is only in less predictable cases that this issue will arise. This presents a real challenge if one is trying to legislate for all eventualities. It is for similar reasons that we would resist attempts to define, or provide guidelines as to, the public interest. Again, this might be possible if the inquiry were limited to particular subject matter (for example, one might ask whether the public interest justified the release of information relating to defence or national security), but adequately defining the public interest as a generality, or what constitutes a matter of public interest, is not possible. Different factors will weigh in the balance, and to different extents, depending on context.
2.245 Whether a communication was, or was intended as, a contribution to a matter of public interest is a matter that must be considered by the court, but a finding either way on the matter of public interest is not determinative of the reasonable excuse element. As a number of consultees rightly noted, there are a complex range of factors to be considered in any situation, and it is not appropriate that one factor (such as intended contribution to a matter of public interest) will always trump another. No doubt the public interest is important, and people should be able to contribute to matters of public interest (even if their contribution might be appear to some to be unsophisticated), but its importance is not such as to override all other interests. One particularly good example of this is the balancing exercise that has to be undertaken when assessing competing Article 8 (right to privacy) and Article 10 (freedom of expression) claims.
2.246 It is for this reason that we would not recommend that the public interest form an explicit element of the offence. Were that the case, to establish guilt the prosecution would have to prove lack of a public interest in a communication: unless the jury or magistrate were sure that there was no public interest in the statement, the prosecution would fail. This is not only difficult to achieve in the absence of a clear definition of the public interest, but it is also not desirable. For the reasons we note above, there may well be an intention to contribute to a matter of public interest, but that does not mean that all other considerations (of the rights of others, say) are to be cast aside.
2.247 However, there are two other important protections that limit the scope of this offence in respect of contributions to matters of public interest. First, to be guilty, the sender must have intended to cause at least serious distress: this will exclude from scope any well-intended communication (even if it might be distressing to some). Second, we are recommending that the press be explicitly exempted from the scope of the offence.
2.248 We accept that there will be difficult cases, as identified by Jacob Rowbottom. Importantly, however, we consider that the potential for difficult cases is mitigated in large part by the move away from an offence based on awareness of a risk of harm.
2.249 Further, because the offence requires proof beyond reasonable doubt that the excuse was not reasonable, it is our view that marginal cases will tend to be decided in favour of the defendant. The contention by ARTICLE 19 that this offence means that freedom is constrained to what a judge considers to be reasonable or in the public interest is, we submit, overstated, given that the burden of proof results in a de facto presumption that the communication was sent with reasonable excuse. The court must be sure (ie beyond reasonable doubt) that it was not.
2.250 We agree with Jacob Rowbottom that there is a risk that, where the other elements of the offence are made out, it may be more difficult to for the defence to argue that the communication was reasonable. Nonetheless, we do not agree that there might never be a reasonable excuse where harm was intended (as suggested by Kingsley Napley). The example cited in the consultation paper, of the parents of a drugdependent child attempting to persuade her to stop taking drugs, is just such an instance where culpability seems minimal despite the intention to cause serious distress.171
2.251 We therefore remain of the view that requiring proof that the defendant lacked a reasonable excuse is a vital element of the recommended offence.
2.252 Our position in the consultation paper was that we should adopt the terms of the Malicious Communications Act 1988 in order to ensure that specific modes of communication were not omitted from the scope of the offence. The offence brings within scope any “letter, electronic communication, or article”. As we stated in the consultation paper:172
Under section 1 MCA 1988, the conduct element covers both electronic communications (such as communications sent via the internet) and non-electronic communications. In our provisional view, the proposed new offence should be aligned with the section 1 MCA offence in this regard. The proposed new offence thereby avoids the pitfalls of the section 127 CA offence in criminalising only certain forms of electronic communication. It also differs from the section 22 HDCA 2015 offence in that it is not limited to digital communications and avoids the problems of a “technologically specific” approach.
However, in another respect, the proposed offence expands on the forms of communication covered under section 1 MCA 1988. As we explain in Chapter 2, the offence under section 1 MCA 1988 requires that the defendant sends a communication of the proscribed character to an intended recipient. Our proposed offence has no such requirement. It would cover posts on social media, such as a publicly available Tweet or a comment below an online newspaper article, both of which have no intended recipient and are accessible to the world at large. Such communications may technically be covered by the word “sent” but, for the sake of clarity, we have added the word “posted”.
2.253 For the avoidance of doubt, “article” is to be understood in the sense of “object” rather than, say, a newspaper or journal article. This was the cause of some confusion in the consultation responses, and we ought to have clarified this position.
2.254 It is also important to clarify that “posted” is intended to cover differing forms of online communication rather than traditional posting of letters or parcels via the mail. The offence is “technology neutral”; in our view, letters and parcels via traditional mail would be in scope of the offence due to the term “sent”, in the same way they are covered currently by the MCA 1988.
2.255 One of the key benefits of framing the offence as technologically neutrally as possible (as we believe we have) is that it means that the offence will not be rendered obsolete in the next five to ten years as forms of communication develop.
2.256 We also note that, where a communication was sent or posted from a device to a social media platform, but was not made visible by that platform (perhaps because of preventative algorithms), it could be impossible for the offence to be made out because the prosecution would have to prove that there was a likely audience who was at a real and substantial risk of seeing the message. It might be that no one was at a real or substantial risk of seeing the communication (ie the likely audience was nobody).
2.257 We recommend that section 127(1) of the Communications Act 2003 and the Malicious Communications Act 1988 be replaced with a new offence with the following elements:
(1) the defendant sent or posted a communication that was likely to cause harm to a likely audience;
(2) in sending or posting the communication, the defendant intended to cause harm to a likely audience; and
(3) the defendant sent or posted the communication without reasonable excuse.
(4) For the purposes of this offence:
(a) a communication is a letter, article, or electronic communication;
(b) a likely audience is someone who, at the point at which the communication was sent or posted by the defendant, was likely to see, hear, or otherwise encounter it; and
(c) harm is psychological harm, amounting to at least serious distress.
(5) When deciding whether the communication was likely to cause harm to a likely audience, the court must have regard to the context in which the communication was sent or posted, including the characteristics of a likely audience.
(6) When deciding whether the defendant had a reasonable excuse for sending or posting the communication, the court must have regard to whether the communication was, or was meant as, a contribution to a matter of public interest.
2.258 We recommend that “likely” be defined as “a real or substantial risk”.
2.259 We recommend that the offence be an either-way offence, with a maximum sentence comparable to the existing Malicious Communications Act 1988.
2.260 As we set out in the consultation paper,173 given the relative seriousness of the recommended communications offence (which is low when compared with other offences normally justifying extradition to the UK), we were of the view that a provision penalising acts committed outside of the jurisdiction of England and Wales would be disproportionate. The offence is also one that criminalises conduct, rather than result.
To that end, the conduct that should be captured ought to be wrongful conduct committed within the jurisdiction. We remain of the view that wholesale extension of jurisdiction is not appropriate. However, as we detail below, we believe that a modest extension of jurisdiction could combat a failing that arises with the existing communications offences, which we do not wish to replicate with our recommended offence.
2.261 We asked a consultation question concerning jurisdiction and extra-territoriality:174
Do consultees agree that the offence should not be of extra-territorial application?
2.262 There was strong support for our view that any offence should be limited to conduct within the jurisdiction.
2.263 The Crown Prosecution Service noted the potentially disproportionate work that could arise if the offence were of wholesale extraterritorial application:175
This is a reasonable proposal, given the extradition of people from overseas responsible for harmful communications received or viewed in England and Wales could involve a disproportionate scope and amount of procedural work. There could be diplomatic implications. For example, the USA has very strong protections for freedom of expression under the First Amendment.
2.264 ARTICLE 19 noted concerns with the principle and practical application of extraterritoriality:176
In our view, the extra-territorial application of such an offence would lead to other countries applying their unduly vague laws beyond their borders, which would have a significant chilling effect on freedom of expression. Moreover, it would be practically very difficult to prosecute individuals based overseas.
2.265 However, a number of consultees expressed their concern with our proposal to limit the offence’s territorial application. 177They outlined the various types of online abuse that are perpetrated by people outside of the jurisdiction and questioned the practical impact that limited jurisdiction would have on any recommended offence.
2.266 A number of consultees conceded that given the practical concerns with any extraterritorial application of an offence, some compromise position could be sought whereby people who are only briefly outside of the jurisdiction are not able to escape responsibility.178 They pointed to various recent examples of this type of behaviour, including examples where a person habitually resident in England and Wales sent tweets while on holiday in Australia and, due to the territorial limitations of the communications offences, had not committed an offence in England and Wales.179 Consultees noted that measures to address similar behaviour form part of the recent Domestic Abuse Act 2021.180
2.267 We accept that any attempt to extend the offence to communications sent from outside the jurisdiction with a likely audience within the jurisdiction would present both practical and principled difficulties. However, we appreciate that there may be limited scope to expand the application of the offence to the extent that any person ordinarily subject to it cannot evade liability by sending a communication while briefly outside the jurisdiction. It is worth considering the decision of R v Smith (No 4). 181In that case, the offence of obtaining services by deception was committed in circumstances where the obtaining had taken place abroad, but a substantial part of the deception had occurred in England. The Court held that not to extend the application of the offence in the circumstances would be an unduly strict interpretation of the doctrine of extraterritoriality. However, central to this reasoning was that different elements of the offence had been committed in different jurisdictions. Given that our recommended offence is complete at the point of sending or posting, the test articulated in Smith does not apply.
2.268 Our recommended offence, similar to the current offences in the MCA 1988 and CA 2003, does not require the recipient (or in the terms of the offence, the “likely audience”) of a communication to be within the jurisdiction. However, any extension of jurisdiction requires consideration of the potential application of an offence where both defendant and likely audience are entirely out of the jurisdiction of England and Wales.
2.269 At this juncture, it is important to note that there is a wide array of relatively serious offences in the criminal law of England and Wales with extraterritorial application.182 Many of these are more serious offences than a communications offence: for example, the offences of murder and manslaughter pursuant to the Suppression of Terrorism Act 1978, a number of terrorism-related offences under the Terrorism Act 2000 and Terrorism Act 2006, and the offences in sections 1 to 3A of the Female Genital Mutilation Act 2003 may all be prosecuted in England and Wales against United Kingdom nationals or habitual residents where the conduct occurs outside of the jurisdiction without any requirement that the conduct is an offence in the jurisdiction where it takes place. Some forms of extraterritoriality use a “local law” requirement - that the impugned act is an offence under the law of the country where it takes place.183
2.270 Offences more analogous to communications offences that are subject to extraterritorial application are the offences of putting people in fear of violence and stalking involving fear of violence or serious alarm or distress in the Protection from Harassment Act 1997184 and the offence of controlling or coercive behaviour in an intimate or family relationship contrary to section 76 of the Serious Crime Act 2015.185 We note that the nature of these offences, similar to our recommended offence, is such that it may be difficult to establish that the impugned conduct matched an existing offence in a different jurisdiction, and that to impose a “local law” requirement may render any extraterritoriality practically unenforceable.
2.271 Further, in line with the observations made by the Government in relation to the Domestic Abuse Bill, it is generally appropriate for criminal offending to be dealt with by the criminal justice system of the state where an offence occurs. 186A prosecution within England and Wales would only take place where the accused person is present in the jurisdiction, there is sufficient evidence to provide a realistic prospect of conviction and it is in the public interest to prosecute.187 Similarly, the Code for Crown Prosecutors notes that decisions in prosecutions involving conduct across jurisdictions will require careful balancing of a range of factors and sets out that the final decision in each case will depend on the circumstances of each case which will: “...weigh heavily on whether there is enough evidence to prosecute and whether it would be in the public interest to do so.”188
2.272 In our view, to ensure that the offence cannot be avoided by those to whom it would ordinarily apply (namely, people habitually resident in England and Wales) by sending or posting communications while temporarily outside the jurisdiction, the offence should apply to communications sent or posted either within England and Wales or outside the jurisdiction by a person habitually resident in the jurisdiction. For completeness, we reiterate that any prosecution arising from a communication sent or posted abroad by a person habitually resident in England and Wales would only take place where there is sufficient evidence to provide a realistic prospect of conviction and it is in the public interest to prosecute.
2.273 This approach ensures that the concerns raised in the consultation paper about proportionality,189 and those noted by consultees about freedom of expression and comity, are respected: the offence will never apply to those who are not either present in, or habitually resident in England and Wales. The offence will extend only to those who are ordinarily subject to the criminal law of England and Wales. We believe that this modest extension of the application of the offence, in line with the approach adopted for the offences in the Protection from Harassment Act 1997 and Serious Crime Act 2015 in the Domestic Abuse Act 2021, will ensure that a current gap in enforcement that exists with the communications offences will not be replicated with our recommended offence.
2.274 We recommend that the harm-based communications offence applies to communications that are sent or posted in England and Wales, or that are sent or posted by a person habitually resident in England and Wales.
3.1 In our consultation paper we set out proposals to reform section 127(2) of the Communications Act 2003 (“CA 2003”). These targeted knowingly false communications and hoax calls to the emergency services. We also asked a separate consultation question about the potential for addressing threatening communications with a specific offence. This chapter addresses both of these topics.
3.2 The criminal law has addressed the most serious forms of false communications for some time. One example is the law of fraud that, among other things, sanctions false representations. 190However, as we set out in the consultation paper, and as many of our consultees confirmed, a specific offence addressing knowingly false communications will bridge an undesirable gap in the criminal law.
3.3 It is important to note at the outset that the proposals were not intended to combat what some consultees described in their responses as “fake news”. While some instances of deliberate sharing of false information may fall within the scope of the offences, our view is that the criminal law is not the appropriate mechanism to regulate false content online more broadly. A substantial amount of “fake news” is shared unwittingly. As a result, this is a matter, appropriately, addressed in some detail in the UK Government response to the consultation on the Online Harms White Paper in the context of non-criminal regulation.191
3.4 The role of the criminal law in addressing persistent communications was also part of our consultation. We recognised the harm that persistent communications can cause. In line with consultees’ responses, we acknowledge the serious harm of persistent hoax calls to public services and outline an appropriately directed offence to address them.
3.5 Regarding threats, certain types of threats are already specifically criminalised. For example, threats to kill under section 16 of the Offences Against the Person Act 1861. We recognised that in reforming the existing communications offences there may be a labelling problem where the general communications offence addresses psychological harm. After considering the ramifications of our recommendations for reform of the communications offences and consultees’ strongly held views on the matter, we believe a specific offence covering threatening communications is appropriate.
3.6 Dealing first with false and persistent communications and then turning to threatening communications, this chapter examines the current law, the provisional proposals for reform and consultees’ responses before recommending specific offences to address each topic.
3.7 We conclude this chapter with a recommendation that is unrelated to section 127 itself but, as with the other behaviours discussed in this chapter, is a particular problem that requires a solution beyond the general harm-based offence. The recommendation relates to the problem of people intentionally sending flashing images to known sufferers of epilepsy. This is a unique problem in the sense that physical harm is caused directly (rather than consequentially, such as communications that encourage self-harm) upon seeing a communication. There is currently no specific offence of sending flashing images in our criminal law, and we could not guarantee that all instances of the behaviour could be prosecuted under the Offences Against the Person Act 1861.3 We did not consult on a specific form of criminal offence for this behaviour, and therefore make no recommendations as to the form of the offence, but we do recommend that the government considers introducing a specific offence to address this reprehensible behaviour.
3.8 Our proposals for reform of section 127(2) CA 2003 were contained in Chapter 6 of the consultation paper.4 We proposed a new offence to address knowingly false communications which was, with the harm-based offence in Chapter 2 above, to form a coherent set of offences to criminalise communications that are likely to cause harm and knowingly false harmful communications.
3.9 As we noted in our introduction, and in the consultation paper,5 more serious forms of false communications are addressed by other criminal offences. One example is section 2 of the Fraud Act 2006. However, in our view, to repeal the existing offences in sections 127(2)(a) and (b) CA 2003 without any replacement would create an undesirable gap in the law.
3.10 Section 127(2) CA 2003, a summary only offence,6 provides:
A person is guilty of an offence if, for the purpose of causing annoyance, inconvenience or needless anxiety to another, he—
(a) sends by means of a public electronic communications network, a message that he knows to be false;
For a discussion of the potential application of the Offences Against the Person Act 1861 see Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 4.123 to 131,
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.7.
Communications Act 2003, s 127(3).
(b) causes such a message to be sent; or
(c) persistently makes use of a public electronic communications network.
3.11 As we set out in the consultation paper, in our view the behaviour targeted by sections 127(2)(a) and (b) warrants separate treatment to that targeted by section 127(2)(c). We believe the former provisions, covering false communications, are best addressed by an offence targeting knowingly false communications. This complements the general harm-based communications offence. The “persistent use” provision, in our view, is best addressed by way of a specific offence prohibiting hoax calls to the emergency services. This will be addressed in Part 2 below.
3.12 Section 127(2) CA 2003 applies to some false communications. Namely, those where the defendant:
(1) sends over a public communications network;
(2) for the purpose of causing annoyance, inconvenience or needless anxiety to another; and
(3) that they know to be false.
3.13 While this provision has the capacity to apply to knowingly false communications that are less serious than the provisions in the Fraud Act 2006, the offence in section 127(2) is infrequently prosecuted. Further, while the provisions in the Fraud Act 2006 may address false communications where there is financial gain or loss, 192there is scope for an offence that addresses false communications more broadly. However, we expressed concern with the existing thresholds of “annoyance” and “inconvenience” in section 127(2) CA 2003.
3.14 By way of reform, we proposed a new offence to replace sections 127(2)(a) and (b), addressing knowingly false communications with a summary only offence with the following elements:
(1) The defendant sent or posted a communication he or she knew to be false;
(2) In sending the communication, the defendant intended to cause non-trivial emotional, psychological or physical harm to a likely audience; and
(3) The defendant sent or posted the communication without reasonable excuse.
(4) For the purposes of this offence:
(a) a communication is an electronic communication, letter or article (of any description); and
(b) a likely audience is someone who, at the point at which the communication is sent by the defendant, is likely to see, hear or otherwise encounter it.
3.15 The similarities between this offence and the general harm-based communications offence are deliberate; they are intended to complement each other as a coherent set of communications offences.
3.16 We asked a consultation question8 about the proposal to repeal section 127(2)(a) and (b) and replace it with the above new offence. Consultees overwhelmingly expressed strong support for the proposal. Some raised concerns with the issue of knowledge of falsity, and a number discussed ramifications for freedom of expression. These responses and the issues they raise are analysed in detail below.
3.17 A range of consultees expressed explicit support for our analysis of the existing communications offences.9 They noted that our proposals would be able better to address harmful false communications, in particular by adopting a “technology neutral” approach. Consultees strongly supported what they saw as a more coherent approach to harmful false communications than the existing offences. Specific concerns were raised regarding freedom of expression and the difficulty of determining and/or sanctioning “falsity”; we will expand on these matters below as we address the specific elements of the offence.
3.18 One other matter that was raised with us in consultation was the extended time limit for prosecution that exists for the current offence in section 127(2) CA 2003. We consider this below and ultimately recommend that extended time limit be retained.
3.19 The offence requires a defendant to send a communication he or she knows to be false, where a communication is a letter, electronic communication, or article (of any description). The mental or fault element, knowledge of falsity, is matched with a conduct element. On this point, the proposed offence does not differ from the existing offence in section 127(2) CA 2003: the defendant must know rather than believe the communication is false. Hence, as relates to the conduct element, the communication must actually be false.
3.20 Our analysis of the limiting “public electronic communications network” requirement in the consultation paper was overwhelmingly supported by consultees. As we noted, this imports an arbitrary distinction between potentially equally harmful modes of communication. This problem is not lessened in the case of knowingly false
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 18.
Consultation Responses: Bar Council of England and Wales, Alan Turing Institute, Magistrates Association, Women’s Aid, Justices’ Legal Advisers and Court Officers’ Service (formerly the Justices’ Clerks’ Society), S Rowe, English PEN.
communications, which are no less problematic in virtue of being sent by Bluetooth or over a work intranet than over the internet, for example.193
3.21 We posed a specific consultation question on the conduct element:194
We provisionally propose that the conduct element of the false communications offence should be that the defendant sent a false communication, where a communication is a letter, electronic communication, or article (of any description).
Do consultees agree?
3.22 There was extremely strong support for this proposal across the spectrum of consultees. A number of consultees noted an oversight on our part in that the provisional proposal only discussed communications which were “sent” rather than those “sent or posted”. For the reasons set out in Chapter 2 in the context of the harmbased offence, we believe it is appropriate for any offence to address communications which were “sent or posted”.
3.23 The Association of Police and Crime Commissioners explicitly supported the removal of a reference to a “public electronic communications network” in any new offence:195
We agree with this proposed approach in terms of not replicating the “public electronic communications network” in the relevant subsection, and agree that false and persistent messages sent for malign purposes are not necessarily any less problematic for being sent via non-public communication networks, such as Bluetooth or a work intranet.
3.24 Samuel Rowe agreed, and discussed the matter of Chabloz v Crown Prosecution Service196 in considering what might amount to “sending”:197
In terms of the act of sending, consideration should be given to the recent case of R (Chabloz) v Crown Prosecution Service. In that case, the High Court held that the offence under section 127(1)(a) was complete with the insertion of a hyperlink that linked to a communication of the proscribed character. The conduct element under section 127(1)(a) and section 127(2)(a) and (b) are indistinct as regards the act of sending. The decision in Chabloz has been criticised on the basis that it misapprehends the nature of hyperlinking and has caused already widely cast offences to become even wider. I suggest that the Law Commission considers carefully the ambit of 'sending' an electronic communication, so as to prevent liability arising where an individual has sent a message that includes a hyperlink to impugnable material but isn't necessarily the subject matter of the communication.
Similarly, consideration should be had as to whether creating a webpage or social network group constitutes 'sending' in the digital context. In my view, it should not, as the proposed offence aims to limit the spread of potentially harmful information, and to do otherwise would likely constitute a disproportionate interference with individuals' Article 10 ECHR rights. The offence should therefore not capture information that is published with no intended recipient or recipients.
3.25 English PEN noted the difficulty in proving falsity, and urged a cautious and narrow approach:198
We note that what is ‘false’ is often difficult to ascertain. The Commission will be well aware of the complex precedents and provisions in defamation law regarding ‘substantial truth’ and ‘justification,’ how the Courts determine whether something is fact or opinion, and whether an innuendo is an assertion of fact.
However, we also note that a version of this offence is already in operation at s.127(2) and that prosecutions under this law are for palpably false claims that can be proven to be so. We assume a narrow definition of falsity will continue to be employed by police and the Courts, and the offence will not be used as a way to litigate what should properly be settled through defamation law.
3.26 The Alan Turing Institute emphasised the importance of requiring knowledge of falsity in their response, arguing that literacy programmes and other initiatives should address unknowingly false communications:199
In many online settings, individuals share false information unknowingly because they believe that it is true or have given it insufficient scrutiny to understand its harmful effects. This is unfortunate, and the unintentional spread of false content should be addressed through other initiatives such as media literacy programmes. However, we do not think that the law is an appropriate way of handling such behaviour.
3.27 Demos noted the difficulties with adopting a “truth/falsity” binary in the approach to criminal offences:200
This point highlights that the nature of false communications online is not understood with sufficient granularity in the proposal so far, as it overlooks, for example, another related category of harmful disinformation: misleading but true information used with malign intent. There are many examples of this, such as the sharing online of true reports of e.g. someone who is a migrant engaging in violence, but which is shared in a context or a manner in which it intends to mislead others into viewing migrants as a whole as a threat. As Demos’ report, Warring Songs, into the variety of information operations that occur online highlighted:
[...] focusing on the distinction between true and false content misses that true facts can be presented in ways which are misleading, or in a context where they will be misinterpreted in a particular way that serves the aim of the information operative.
The presumption that a true/false binary is sufficient to capture many of the abuses of false communication in an offence is, therefore, mistaken. Again, however, this point is meant to illustrate the broader point that consideration of truth and falsity is not the only dimension by which this proposal could be made more granular and thus more effective.
3.28 As regards “sending”, the use of “sent or posted” in this offence mirrors the approach taken with the harm-based offence in Chapter 2. As we set out there, in our view the combination of the fault and conduct elements in the harm-based offence (and, here, for the offence addressing knowingly false harmful communications) are sufficient to ensure that any innocuous or innocent use of hyperlinking will not be inadvertently captured.
3.29 We accept the difficulties that may attend discerning falsity in practice, and the potential harm that misrepresentation may give rise to in other communications. However, we note that where communications wilfully misrepresent “true” information, or otherwise communicate without the requisite falsity for this offence, the general harm-based communications offence is relevant.
3.30 Where communications are sent with the intention wilfully to misrepresent and cause serious distress, such that they satisfy the elements of our separately recommended harm-based offence, they can be appropriately criminalised. While these may have to meet the more onerous threshold of harm required by that offence, the distinction lies, in our view, in the differing utility between knowingly false harmful communication and misrepresentation. Moreover, in circumstances where a communication is knowingly false, the justification for interfering with the sender’s right to freedom of expression is stronger than it is for true but misleadingly used communications.201
3.31 In response to consultees who raised concerns that the scope of the proposed offence may be too narrow, we accept that the proposals narrow the scope of liability when compared with section 127(2) CA 2003. However, as we set out in the consultation paper,202 the scope of the existing offence leads to both over and under criminalisation. In our view, though there may be a gap between the existing offence and our recommended one, we believe that our recommended offence ensures that only those false communications properly characterised as criminally culpable will be captured. Against this, we accept that true information can be used in malign ways. In our view, those communications should be dealt with by way of the general harm-based offence outlined in Chapter 2.
3.32 Further, we agree with the comments English PEN made in their consultation response in considering the conduct element. 203This offence is aimed toward (albeit not limited to) “palpably false” claims capable of proof. This is not a novel approach, and the conservative approach adopted by courts militates against concerns that the proposed offence would be used as a way of litigating concerns that should be the subject of defamation action. It also draws out a part of our proposals designed to ensure that criminal sanction is directed appropriately. The aspect of the fault element requiring knowledge of falsity ensures that a difficulty that could otherwise arise (determining “degree” of falsity as an objective matter) does not. As we noted in the consultation paper, the elements in combination are designed to ensure that the offence “catches only communications which lack social utility.”204
3.33 As we set out in the consultation paper, the proposed fault element of the offence addressing harmful false communications has two aspects:205
(1) the defendant knew the communication to be false (knowledge of falsity); and
(2) the defendant, in sending the communication, intended to harm a likely audience, where harm is defined as any non-trivial emotional, psychological or physical harm.
3.34 The existing provisions in sections 127(2)(a) and (b) do not cover messages that the sender believes to be true. As we set out by way of example, where a person sincerely believes that vaccination against various illnesses is bad for a child’s health, and sends communications to this effect, the provisions would not apply. 206In line with the Government’s advice on introducing or amending criminal offences, and to ensure that the offence reflects a sufficient degree of culpability on the part of defendants, we proposed that the new offence replicate the requirement of knowledge of falsity.207
3.35 We set out the various other initiatives and avenues of addressing sincerely believed false communications in the consultation paper, and note that “misinformation” along these lines is given specific consideration in the UK Government response to the Online Harms White Paper.208 As we argued in proposing the new offence, where false information is communicated with knowledge of falsity, coupled with some “malign purpose”, there may be a role for the criminal law to play.
3.36 The proposal seeks to ensure that the criminal law only addresses knowingly false communications in circumstances where there is sufficient harm, but addresses an appropriately broad range of potential harms.209 It also seeks to ensure that where a defendant claims to act for some other ulterior “purpose” they cannot escape liability.
3.37 We asked a specific consultation question on this aspect of the offence:210
We provisionally propose that the mental element of the false communications offence should be:
1) the defendant knew the communication to be false; and
2) the defendant, in sending the message, intended to harm a likely audience, where harm is defined as any non-trivial emotional, psychological, or physical harm.
Do consultees agree?
3.38 The Crown Prosecution Service agreed, noting the differences from the existing communications offences that arguably make the proposal more onerous to prove:211
We agree. The proposed false communications offence would require intent, as required by existing law under section 1 [Malicious Communications Act] 1988 [(“MCA 1988”)] and section 127 CA 2003. The new offence would not cover 'belief' that the communication was false (as under section 1 MCA 1988). Section 127 CA 2003 also requires that the defendant knew the communication was false, rather than believed it to be so.
3.39 Fix the Glitch supported the proposed approach. They emphasised that misinformation and disinformation can disproportionately impact certain users more than others, referencing the gendered and racialised components of some harmful communications:212
We agree that the mental element of the false communications offence should include the two elements above, covering both the deliberate sharing of disinformation and intention to cause harm. We would also like to draw attention to the ways in which misinformation can cause harm and the ways in which disinformation and misinformation have a gendered and racialised component, affecting certain users more than others.
3.40 Community Security Trust disagreed, noting the potential loophole that may be created which may not address common antisemitic abuse. They provided evidence of the types of behaviours they are concerned about and urged any proposal to focus on the harm that a victim suffers rather than a perpetrator’s motivation or knowledge:213
The consultation explains in point 6.60 that 'We do not, however, propose to cover communications that the defendant believes to be true - no matter how dangerous those communications may be. We recognise that misinformation and ‘fake news’ are serious social problems, but they lie beyond our Terms of Reference'.
This has the potential to leave a loophole whereby common and accepted forms of antisemitic discourse will not be considered as offences if they otherwise fall under the terms of the proposed changes to online communications law, whereby Jewish individuals or organisations are being targeted with antisemitic abuse.
Examples include common antisemitic canards and conspiratorial narratives around 'Jewish Power', which claim that Jews somehow rule and manipulate the world via assorted mechanisms of global control.
Though this is not illegal per se, CST has published evidence of many examples whereby Jewish individuals or organisations are harassed and targeted online in an antisemitic manner by offenders employing these themes - whether dressed up as a 'Rothschild Conspiracy Theory' or not.
Another example is the denial or distortion of the Holocaust. Though, in the UK Holocaust Denial is not illegal, it can certainly be employed to harass and target members of the UK Jewish community in an antisemitic manner.
The offender may truly believe in their misguided notions, but the onus shouldn't be on whether or not they believe in what they are saying or not. It should be on the impact that their 'false communication' has on the victim, especially when employed in a hateful manner that targets a specific (or several) protected characteristics.
3.41 Samuel Rowe disagreed, arguing that in order to retain important flexibility in application, the proposed offence should also capture wilful blindness:
I disagree that the defendant should have to know the communication to be false, applying a wholly subjective test to the defendant's knowledge.
In DPP v Collins, Lord Bingham held that the mental element of section 127(2) requires "proof of an unlawful purpose and a degree of knowledge" (at p 10). In my view, the proposed offence should capture both actual knowledge and wilful blindness. This is because the law should be flexible enough to punish the most severe cases of recklessness and because establishing actual knowledge will likely be extremely difficult to establish to the requisite evidential burden.
However, I do not go as far to recommend that constructive knowledge be sufficient to establish the mental element of the proposed offence. Constructive knowledge will very rarely establish criminal liability and with this offence it should be no different.
3.42 English PEN noted their reservations about the offence itself (referencing their other answers which pointed to other existing legal mechanisms for addressing mis- and dis- information). They did note that, should the proposals proceed, they support the wording of the provisional proposals, in particular knowledge of falsity:214
We question the need for this offence.. .However, if a new offence is to be introduced, we support the proposed wording, for the reasons set out in paragraphs 6.42 and 6.46.
Knowledge of falsity is a crucial addition that will limit the scope of the offence in a way that will protect individual rights to express opinions and share information.
3.43 Professor Alisdair Gillespie noted that while extending the fault element to include recklessness would be controversial, it could be a way of addressing harmful further dissemination (eg retweets), and also provided an example to query whether a slightly different formulation of the required purpose element is more appropriate:215
I can see the logic of this, but it is challenging of further dissemination (e.g. retweets). Knowledge is a high threshold to clear, and some offences do have an equivalent of recklessness (‘ought to have known’). Extending it to include recklessness would be controversial, but there is a danger that harmful content could be sent through retweets.an alternative (and possible half-way house) is to say that one of the purposes must be to cause non-trivial emotional, psychological or physical harm. That is used elsewhere in the law.
3.44 Full Fact provided a detailed consultation response that grappled with the potentially problematic nature of determining falsity. 216They noted that in their experience, misinformation is “often deliberately designed to be not false but to create a false impression” and that it “is often simple to manipulate a false claim into a true claim that is in effect misleading.” To this end they note the challenges of proof that will exist for any offence dealing with falsity.
3.45 The Justices’ Legal Advisers and Court Officers’ Service (formerly the Justices’ Clerks’ Society) argued that “non-trivial” may be a difficult concept to define. Further, they raised a concern with the definition of harm. They referred to the serious
consequences that the making of malicious complaints can have and questioned whether the proposals could adequately address this:217
Non-trivial is hard to define and we wonder whether a clearer definition could be sought.
We are also concerned that the definition of harm is too restricted and would decriminalise some undesirable behaviour which is currently criminal under the Malicious Communications Act and the Communications Act - e.g. making malicious complaints with a view to causing purely material or reputational harm, such as public calumny, loss of position (paid or voluntary), or exposing the victim to criminal investigation. While some offences could be prosecuted as fraud, frequently such offences will not satisfy the requirement of intent to cause loss of property. With such offences the emotional harm is often secondary to the material harm, and the material harm is often what the offender intended.
Misinformation and the criminal law
3.46 Community Security Trust brought to our attention the antisemitic abuse that is often predicated on or perpetuated by false information.218 It is true that the proposed formulation would only criminalise a subset of harmful false communications. However, if lower thresholds of knowledge or intention were adopted, it is inevitable that a broader range of behaviour would be captured. While that may address a greater range of culpable behaviour, it also carries the real risk that non-culpable behaviour would be criminalised. Given that our concerns with the existing offence largely related to its overly broad scope, 219we think that a new offence should be more appropriately constrained and directed. However, there are other potential offences that can apply to abuse predicated on or perpetuated by false information. 220We also deal with persistent and other “low-level” communications in Part 2 below.
3.47 We also note that the concerns raised about the use of “true” information in a harmful way (which many consultees referred to as “misinformation” as opposed to “disinformation”) may well be subject to our recommended harm-based offence. In circumstances that a person sends a communication that contains “misinformation” intending to cause harm to a likely audience, where that communication is likely to cause harm to the relevant threshold, they will be liable. Our consideration of knowingly false communications seek to address the related, but different, phenomenon of the communication of harmful disinformation or otherwise knowingly false communications.
3.48 The most serious harmful false communications may in some circumstances also be subject to the stirring up hatred offences found in Part III of the Public Order Act 1986. These offences are discussed at length in our consultation paper in a separate project on hate crime. 221While the provisions are complex, they provide alternative ways to ensure that a range of behaviour, including extremely serious intentional online communication where false information is used to stir up racial or religious hatred or hatred on the basis of sexual orientation, is subject to the criminal law.222 We have provisionally proposed to remove the requirement that the words or behaviour be “threatening, abusive or insulting” where intention to stir up hatred can be proved.223 This would allow the deliberate spreading of untruths about particular groups to be specifically addressed. The examples we consider in the consultation paper include antisemitic tropes such as the “blood libel” or the Protocols of the Elders of Zion.224 These offences are much more serious than the summary offence in section 127(2) CA 2003, and, in our view, can more appropriately reflect the pernicious harm that can be caused by knowingly communicating false information to stir up racial hatred.
3.49 Other more serious criminal offences that are predicated on false information include the offence of making a bomb hoax in section 51 of the Criminal Law Act 1977, which has a maximum penalty of seven years imprisonment.225 Further, section 38 of the Public Order Act 1986 criminalises contamination of or interference with goods with the intention of causing public alarm or anxiety. The offence explicitly includes the circumstance where a defendant makes it appear that goods have been contaminated or interfered with, and has a maximum penalty of up to 10 years imprisonment. 226The substantial penalties attached to these offences, and the seriousness of the conduct they seek to prevent, can be contrasted with the more modest aim of an offence targeting knowingly false harmful communications.
3.50 Further, our general communications offence recommended in Chapter 2: above would also be a more serious “either-way” offence than the summary false communications offence that could be used in circumstances where a person sends a communication with the intention of causing harm to a likely audience. This could apply to single communications of the type that Community Security Trust describe where they are sent with the requisite intent; it would also provide a greater maximum penalty given we recommend that offence is triable “either-way” as compared with our recommendation that the false communications offence be tried summarily.
3.51 Some consultees raised the varying harms that can be caused by false communications in their responses and explored the benefits of an offence with a broader scope.227 However, as for the general communications offence, it is important to recognise that any communications offence must take care to not attempt to address harms that are too diffuse or too remote, or could be better addressed by other criminal offences. It is not clear how potential remoteness concerns could be adequately contained if the proposals were expanded. The varying types of harm that communications can cause were the subject of extensive analysis in the consultation paper.228 We acknowledge the varied types of harm that may form part of the “knock-on effect”229 flowing from the initial harm caused by a harmful communication.
However, in order to ensure that our proposals do not disproportionately expand the scope and application of communication offences, it is important to ensure that any harms captured have a sufficient causal connection with the defendant’s conduct.230
3.52 As we set out in the consultation paper in relation to the existing communications offences, the risk of over-criminalisation arising from vagueness, 231as well as the unsatisfactory targeting and labelling that occurs when these offences are applied in different contexts (particularly relating to online communications),232 make us cautious in recommending any offence that does not have a clearly defined scope.
Knowledge of falsity and intention to cause harm
3.53 As set out above a number of consultees noted that the proposed threshold will make the offence quite difficult to prove: both knowledge of falsity and intention will make “borderline” cases particularly difficult. 233However, the inverse risks have been laid bare by consultees’ responses to the lower thresholds proposed for the general communications offence. We were grateful for the various responses 234that explored the different potential options for this aspect of the offence, including recklessness or requiring that the defendant “ought to have known” the communication was false.235 These lower thresholds would broaden the scope of the offence. As we set out in the consultation paper, 236we are anxious to ensure that any new offence does not have a scope that reflects an extremely low level of culpability. While we accept that these lower thresholds would allow a broader range of communications to be captured, in combination with the threshold of harm in the second aspect of the fault element we are not convinced it would represent a proportionate interference with freedom of expression.
3.54 Regarding the second aspect of the fault element, intending to cause harm to a likely audience, other consultees including Big Brother Watch argued that, despite the requirement of knowledge of falsity, requiring intention to cause “non-trivial emotional harm” does not ensure sufficient culpability.237 While there was extremely strong support for a replacement offence, there was a variety of views as to how best to formulate a replacement. The variety of views among stakeholders demonstrates the difficult task at hand: ensuring that the criminal offences we propose do not disproportionately interfere with freedom of expression and appropriately address the relevant harm. The latter aspect of the fault element, requiring an intention to cause harm, is also set at a higher threshold than the existing offence in the CA 2003.238 Further, for the same reasons we provided in Chapter 2 in relation to the general harm-based offence we believe that it is appropriate to remove reference to “emotional” harm.239
3.55 In isolation, simply requiring knowledge of falsity or an intention to cause harm to the relevant threshold may not direct the offence appropriately. Similarly, on its own, the threshold of “non-trivial” psychological or physical harm appears to be a low threshold. Importantly the two aspects of the fault element will not operate in isolation. They will be taken together. In our view this combination of the two aspects of the fault element will ensure that criminally culpable harmful false communications are addressed in a proportionate way.
3.56 Despite the various issues discussed above, there was extremely strong consultee support for the provisionally-proposed fault element of the false communications offence. The requirement of knowledge of falsity, coupled with the requirement of intention to harm, largely addressed concerns that consultees expressed with the lower thresholds in the general communications offence explored in earlier questions. While we accept that false information can cause different types of harm, as we set out above, there are various ways to address this behaviour that would avoid an overly broad offence. Ultimately, we believe that the proposed fault element will ensure that the offence only captures criminally culpable false communications.
3.57 As for the general communications offence, we provisionally proposed that the offence addressing harmful false communications should have an additional safeguard for freedom of expression. While, as for the general offence, there was some confusion over the exact nature of the safeguard, the proposal was that the prosecution be required to prove, beyond a reasonable doubt, that any communication was sent without reasonable excuse. Some consultees characterised this as a positive “defence”, that defendants would need to prove themselves: this was not the case. Rather, as it would be an element of the offence, if the prosecution could not prove beyond reasonable doubt a lack of reasonable excuse, the defendant would be acquitted.
3.58 We asked a specific consultation question on this point:240
We provisionally propose that the false communications offence should include a requirement that the communication was sent without reasonable excuse. Do consultees agree?
3.59 The Criminal Law Solicitors’ Association supported the proposal and provided a positive assessment: “This is not contentious and is appropriate and proportionate.”241
3.60 Samuel Rowe noted the importance of the “reasonable excuse” safeguard for freedom of expression, explicitly agreeing that it should not be a defence but instead be for the prosecution to disprove:242
I agree as it is not hard to comprehend circumstances where a defendant will be able to establish a reasonable excuse.
Therefore, in order prevent over criminalisation, as well as unlawful interferences with individuals' Article 10 ECHR rights, it should be possible for individuals to have a reasonable excuse for sharing harmful, false information.
I also agree that the reasonable excuse should not form a defence, since the evidential burden for proving the message was sent without a reasonable excuse should rest with the prosecution.
3.61 The LGB Alliance disagreed with the proposal. They argued that it would be inappropriately censorious to enact a law whereby a communication that upsets another would be criminal without “reasonable excuse” on behalf of the sender:243
The notion that communications that may upset someone must have a ‘reasonable excuse’ amounts to censorship.
3.62 English PEN repeated their position that while they are not convinced of the need for a false communications offence, that if it is to proceed, the “reasonable excuse” protection is essential to ensure an appropriately narrow scope:244
If a new offence is to be introduced on these terms, then a ‘reasonable excuse’ requirement is essential for limiting the scope of the law.
If sub-section (3) were absent, the proposed law would amount to a disproportional interference with Article 10.
3.63 Refuge raised concerns with potential manipulation of any “reasonable excuse” by perpetrators of abuse:245
...whilst we do not oppose the inclusion of a without reasonable excuse requirement in its entirety, we are concerned that unless very tightly defined, this requirement could be manipulated by perpetrators who attempt to justify their abuse and control as an act of love and care.
3.64 Some consultees expressed a degree of scepticism over the utility of a “reasonable excuse” safeguard in circumstances where a defendant has intended to cause harm and has knowledge of the falsity of a communication. This point was, however, balanced against those who acknowledged this potential argument and emphasised the importance of the safeguard regardless. These stakeholders generally noted that affording greater protection for expression and communication than may be required by a “threshold” or “floor” of Article 10 is not a strong argument against including the safeguard.
3.65 We acknowledge that some consultees felt that even with the safeguards built into the proposals, there would be an unacceptable interference with the ability to communicate on contested issues. However, these views were by and large expressed by a minority of consultees who consistently opposed communications offences of any kind. In our view, the various safeguards for expression (requiring knowledge of falsity, an intention to cause harm and that the prosecution prove the communication was without reasonable excuse) will ensure that only criminally culpable communications can fall within the scope of the offence.
3.66 Despite the potential concerns canvassed above, there was extremely strong consultee support for this proposal which was variously described as “crucial” and “important” as a safeguard for expression.
3.67 We accept that some consultees are of the view that safeguards built into offences are insufficient protection for freedom of expression. As we set out in the Introduction and in Chapter 2 in relation to the harm-based offence,246 it is our view that in addressing criminally harmful communications, the most proportionate way to protect freedom of expression is by ensuring that specific safeguards exist in the offences themselves. This proposal not only requires intention and knowledge of falsity, but also proof beyond reasonable doubt of a lack of reasonable excuse. These safeguards, individually but particularly in combination, provide robust protection for freedom of expression. They combine to target only the communication of knowingly false information with intent to cause harm to another without reasonable excuse: behaviour clearly capable of criminal sanction in line with Article 10.
3.68 Refuge’s concern about abusers potentially manipulating the “reasonable excuse” justification is an important one. Individual elements of offences may well provide opportunities for perpetrators to attempt to evade liability. However, whether in supplementary guidance or in careful interpretation of “reasonable excuse” itself, any perpetrator’s act of abuse would be seen in its context. Further, if false communications sent by perpetrators of domestic abuse formed any pattern of harassment, then the relevant harassment offences would apply irrespective of the elements of the existing or recommended replacement communications offences.
3.69 The strength of support our provisional proposals received across the spectrum of our stakeholders and consultees fortifies us that a new summary offence to replace sections 127(2)(a) and (b) CA 2003 is appropriate. In our view, the existing maximum penalty of six months’ imprisonment247 provides adequate scope to punish the type of behaviour that our recommended offence would target. Finally, we agree with the observation of the Justices’ Legal Advisers and Court Officers’ Service that the extended time limit contained in section 127(5) CA 2003 should be replicated.248 While the ordinary limitation for summary offences is six months,249 this was extended in relation to the CA 2003 by the Criminal Justice and Courts Act 2015. 250As set out in the explanatory notes 251and the fact sheet accompanying the Bill on this point, 252this extension was to allow more time for prosecutions to be brought in response to representations raised with Ministers that the six month time limit “[hampered] police investigations into internet related offences”. 253In our view it is appropriate to mirror this in our recommended summary offence that targets knowingly false harmful communications.
3.70 As set out above, we believe that the conduct and fault elements, and the reasonable excuse safeguard, will ensure that the offence only targets culpable false communications, and has robust protections for freedom of expression.
3.71 We recommend that the existing offences in sections 127(2)(a) and (b) of the Communications Act 2003 be replaced with a new summary offence targeting knowingly false harmful communications with the following elements:
(1) the defendant sent or posted a communication that he or she knew to be false;
(2) in sending the communication, the defendant intended to cause non-trivial psychological or physical harm to a likely audience; and
(3) the defendant sent the communication without reasonable excuse.
(4) For the purpose of this offence:
(a) a communication is a letter, electronic communication or article (of any description); and
(b) a likely audience is someone who, at the point at which the communication was sent by the defendant, was likely to see, hear, or otherwise encounter it.
3.72 We recommend the offence should have a limitation period of three years from the day the offence was committed and six months from the day the prosecution becomes aware of sufficient information to justify proceedings.
3.73 The offence in section 127(2)(c) of the CA 2003 has its origins, in part, in protecting public communications networks and those involved in their operation. As we noted in the consultation paper, a random sample of prosecutions provided to us by the Crown Prosecution Service saw 17 of 26 matters involve repeated hoax or vexatious or trivial calls to emergency services.71 We noted that the interpretation of the “public electronic communications network” means that the application of section 127(2)(c) CA 2003 is incredibly broad. For example, repeatedly sending text messages to a family member to annoy them (whether or not it did annoy them) is a criminal offence. We are not satisfied that this is justifiable: neither the level of fault nor the level of harm warrant
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.11.
criminal sanction.254 As was noted by Warby J in a judgment delivered after the publication of the consultation paper, the object of section 127(2)(c) was:255
...to prohibit the abuse of facilities afforded by a publicly funded network by repeatedly exploiting those facilities to communicate with another for no other purpose than to annoy them, or cause them inconvenience, or needless anxiety.
3.74 We acknowledge that certain types of persistent communication can be harmful and justifiably criminalised. However, we believe that a more targeted offence than section 127(2)(c) CA 2003 is required. As we noted in the consultation paper, persistent communication in the context of domestic abuse can be extremely harmful. Specific offences designed to protect against this behaviour in the context of domestic abuse are better suited to addressing the harm: the powers to make domestic abuse protection orders (breach of which would be a criminal offence) 256and the offence in section 76 of the Serious Crime Act 2015 addressing controlling or coercive behaviour in an intimate or family relationship. 257Further, the general communications offence recommended in Chapter 2: would capture communications likely to cause psychological harm. The context the court must take into account when establishing whether harm was likely would include having regard to whether they form part of a pattern of domestic abuse or are persistent.
3.75 As we noted in the consultation paper, Ofcom, the regulator for communications services, is empowered under section 128 CA 2003 to issue notifications for persistent misuse of electronic communications networks or electronic communications services, enforceable in the civil courts under section 129.258 While the terms of section 128 CA 2003 use the test that we have described as problematic in connection with criminalisation, 259given the gravity and consequences of criminal conviction, it is possible its use in relation to civil remedies does not give rise to the same concerns. Given our terms of reference are explicitly limited to the criminal law we do not propose to analyse the substance of this provision further.
3.76 We posed a specific consultation question on repealing and replacing section 127(2)(c):260
We provisionally propose that section 127(2)(c) should be repealed and replaced with a specific offence to address hoax calls to the emergency services. Do consultees agree?
3.77 Consultees expressed extremely strong support for the proposal and our analysis of the existing provision. Even consultees who are vigorously opposed to other aspects of the project, including ARTICLE 19 and the Free Speech Union, noted their support for an appropriately directed offence of hoax calling.
3.78 Some consultees, while supporting the action on hoax calls, noted the potential gap that may be created in repealing section 127(2)(c) without a more comprehensive replacement. The Crown Prosecution Service noted in particular that while “low-level” behaviour may fall under the auspices of the provision, the harm that results may be substantial. This is a valid point. However, these responses did not engage with how the more comprehensive harm-based communications offence may cover this type of behaviour. Further, as we consider elsewhere in this report (particularly in relation to group harassment), the offences in the Protection from Harassment Act 1997 can address persistent behaviour that may not satisfy the threshold of our recommended “harm-based” communications offence. “Low-level” behaviour, including persistent hoax calling, could well form a “course of conduct” for the purposes of that Act.261
3.79 Some consultees responded disagreeing with the provisional proposals in their entirety.262 They raised their concerns that the proposed reforms may result in overly complex laws, and noted that retaining the existing provisions may be the simplest solution.
3.80 We accept that no law should be unnecessarily complex, and where existing provisions adequately address harmful behaviour, there may be no case for reform. However, these responses (which were in a small minority) do not meaningfully engage with our analysis in the scoping report 263or consultation paper. 264Rather, in our view, which was largely vindicated in consultation, the existing provisions have an overly broad scope. Further, our provisional proposal on this topic was designed specifically with simplification in mind: to recommend a narrower offence that targets specific harmful behaviour.
3.81 The Justices’ Legal Advisers and Court Officers’ Service, formerly the Justices’ Clerks’ Society, agreed with the proposal:265
Yes, we agree that a bespoke offence with a proportionate maximum penalty would be appropriate.
3.82 Policing stakeholders agreed with the proposals, including the National Police Chiefs’ Council. 266The Association of Police and Crime Commissioners agreed that the existing offence is no longer appropriate:267
We support this proposed approach, and agree with the Commission’s concerns that the expansion of the definition of the word “network” has led to the offence now being very broad, at least in theory. We support the move to update the law to reflect the nature of how people communicate now in online spaces.
3.83 ARTICLE 19 agreed, providing an offence was clear and of narrow scope:268
In principle, ARTICLE 19 would be supportive of a specific offence to address hoax calls to the emergency services as long as it is sufficiently clearly and narrowly defined so as to protect freedom of expression.
3.84 The Free Speech Union agreed, referencing the recent decision of Scottow v Crown Prosecution Service, 269which involved disputed interpretations of section 127(2)(c).270
3.85 The Magistrates Association agreed that should section 127(2)(c) CA 2003 be repealed, a specific offence be created to address hoax calls to the emergency services:271
This seems a sensible proposal. The consultation sets out a convincing argument that currently section 127(2)(c) risk criminalising certain behaviour that is not intended to be covered by any offence. However, it is obviously important that if section 127(2)(c) was repealed, there was legislation to cover specific offences, including hoax calls to the emergency services. Other examples given in the consultation paper, can be dealt with elsewhere, including via legislation covering domestic abuse.
3.86 The Crown Prosecution Service noted the potential harm that “low-level” behaviour can cause:
We understand the rationale for the proposal and there is some value to it. However, it is important not to underestimate the impact that other low-level behaviour can have when it is persistent. The views of victim groups could also add valuable insight to consideration of this proposal.
We note that where the harm is more serious such as domestic abuse then targeted laws such as the Domestic Abuse Bill 2020 and [section] 76 of the Serious Crime Act 2015 may be more appropriate.
3.87 We agree that persistent “low-level” behaviour can have a serious impact on individuals. However, as we set out in the consultation paper, we believe that the most serious examples of this behaviour can be better addressed with separate specific offences. We further note that the offences contained in the Protection from Harassment Act 1997 provide flexible and specific offences that deal with “courses of conduct” - which persistent “low-level” behaviour could well amount to. As such, we do not think that broadening any offence to combat hoax calls to the extent that it can capture any “low-level” persistent communications is appropriate.
3.88 The Carnegie UK Trust noted the possible protection that a prohibition on persistent misuse of public networks may afford in future:
As regards section 127(2) Communications Act, the Report notes that the main purpose has been to deal with hoax calls to the emergency services. While we agree that this is a distinct objective that could be dealt with via a specific offence, we also note that under section 128 Ofcom may bring enforcement action against those who persistently misuse electronic networks and services. The majority of these have been related to silent calls. While many of these may arise from automated dialling systems and cold-calling, not all will fall into this category, but could be from, for example, ex-partners or people trying to elicit a reaction for their own amusement. Ofcom notes that silent calls are more annoying and more likely to cause anxiety. The circumstances in which people could persistently misuse public networks using modern communications software and techniques are many, varied and unforeseeable in the future. We invite the Law Commission to consider whether such behaviours would be adequately dealt with under the proposed new regime and emphasise that a focus on new services (e.g Skype, WhatsApp) should not mean that older services that are still much used are forgotten.
3.89 The Bar Council of England and Wales noted the possible applicability of the offence of wasting police time, but went on to note that, if the existing position is unsatisfactory, that the proposed replacement offence would be appropriate:
This proposed law reform is motivated by what the Law Commission identifies in this consultation as the need to legislate for harm-based offences (§6.3). Existing legislation is intended to cater adequately - and comprehensively - for the mischief and misconduct of a person making a hoax call to police. The offence of wasting police time, contrary to section 5(2) of the Criminal Law Act 1967, is a type of public justice offence for which the Crown Prosecution Service has published specific and current guidance^We doubt that the offence specified by section 5(2) is a complete answer to the conduct discussed in the consultation:
(i) The offence is directed at ‘wasteful employment’ rather than the harm caused by the hoax itself - ie public fear and inconvenience. What if a hoax call is made and a public place is evacuated with one tannoy announcement which is shortly afterwards cancelled because the hoax is quickly revealed? The police time wasted may be minimal but the ‘harm’ may be considerable (the Crown Prosecution Service suggests ten (10) hours).
(ii) The offence in section 5(2) only relates to wasting police time. What if a hoax caller rings up a school or a branch of a department store or rings a fire rescue service? The police will almost certainly, but not inevitably, become involved before the hoax is uncovered.
(iii) Section 5(2) is not a bespoke communications offence whereas the existing offence for which section 127(2)(c) provides is and presently covers hoaxes too. So there is already an element of overlap.
We invite the Law Commission to consider this. If a new communications offence is necessary, the Bar Council agrees with the Law Commission’s proposed approach.
3.90 This detailed response was useful in reflecting on the purposes for the provisionally proposed offence targeting hoax calls. However, as the Bar Council noted, simply repealing section 127(2)(c) CA 2003 and relying on the offence of wasting police time may well not adequately protect against the harm that hoax calls to the emergency services causes. Most importantly, it could not protect against hoax calls to emergency services other than police. Further, as the vast majority of consultees noted, this discrete type of harmful communication is amenable to a specific offence. We do note that the final form of any offence should be flexible enough to ensure hoax calls to all emergency services are within its scope.
3.91 In relation to the concerns raised by the Carnegie UK Trust, as we set out at paragraph 3.75 above, given our terms of reference we do not propose to make any recommendations about the powers of Ofcom under section 128 CA 2003. These powers will continue to exist regardless of our recommendations on the criminal offences in section 127 CA 2003.
3.92 In our view the offence should include, so far as possible, hoax calls to the emergency services, however they are made. 272While certain types of particularly harmful or disruptive hoaxes cover much broader types of behaviour (going well beyond “calls”, however they are made), 273this recommendation is concerned with the generally harmful behaviour of hoax calls to the emergency services. We accept that this offence will have a narrower scope than section 127(2)(c) CA 2003: that was a central aspect of our proposal, given what we saw was the unduly broad scope of the existing provision. 274As we have set out above at paragraph 3.78, there are other existing offences that address repeated vexatious or trivial communications where appropriate, including those in the PHA 1997. However, the specifically serious harm posed by hoax calls to the emergency services is amenable to a separate offence.
3.93 In the result, the extremely positive response we received during consultation fortified our view that section 127(2)(c) of the CA 2003 should be repealed and that a specific summary offence addressing hoax calls to the emergency services is appropriate. As for the offence addressing knowingly false communications recommended above, we are of the view that the maximum penalty for this offence should be six months’ imprisonment. This reflects the existing penalty, 275and, in our view, remains an appropriate threshold.
3.94 We recommend that the offence in section 127(2)(c) of the Communications Act 2003 be replaced with a summary offence to address hoax calls to the emergency services.
3.95 The harm-based offence recommended in Chapter 2: would capture threatening communications where the elements of the offence are satisfied. However, the general nature of the harm-based offence is such that there is not explicit recognition of threatening communications, unlike the “menacing” communication aspect of section 127 CA 2003 and section 1(1)(a)(ii) of the Malicious Communications Act 1988, which specifically criminalises threats.
3.96 As we noted in the consultation paper, genuine threats are a specifically harmful type of communication. There are existing provisions that criminalise certain types of threats 276and in our report on Offences Against the Person we specifically recommended that a reformed statute governing offences of violence should include an offence of threatening to kill, cause serious injury or to rape any person, including where the threat is made conditional on the conduct of the person to whom the threat is made or any other fact or event.277
3.97 We use the term “genuine” in relation to threats to reflect a level of seriousness. As we set out in our recent consultation paper on Intimate Image Abuse,278 requiring a defendant to have intended the victim of a threat to fear that the threat would be carried out, or be reckless as to the fact, ensures that only sufficiently culpable behaviour is captured by a threat offence. This behaviour can be contrasted by the facts of Chambers.279 Paul Chambers wrote a Tweet saying “Crap! Robin Hood Airport is closed! You’ve got a week and a bit to get your shit together otherwise I am blowing the airport sky high!!”. He was initially convicted under section 127(1) of the CA 2003, but in allowing his appeal, the High Court held that on an objective assessment the
Tweet was not capable of being properly construed as “menacing”.280 In our view, requiring that the defendant intend or be reckless as to whether the victim of the threat would fear that the defendant would carry out the threat will ensure that only “genuine” threats will be within the scope of the offence. Communications sent without the requisite fault will not be within the scope of the offence.
3.98 Genuinely threatening communications are a particularly harmful subset of harmful communications. They have been subject to scrutiny as a result of the change in nature of our communications during the COVID-19 pandemic - as we discussed in the consultation paper:281
...Matthew Wain was recently convicted for posting a YouTube video in which he said that he hoped the National Health Service (“NHS”) staff at Birmingham City Hospital would “all die of the coronavirus because they would deserve it." He also said, "Not being funny... after what I had done to me yesterday I would bomb the place, to be honest.” As District Judge Briony Clarke observed, part of the wrongful nature of Mr Wain’s conduct was that it was threatening.
3.99 A similar issue has arisen relating to threats made to journalists, which has resulted in the Government formulating a specific action plan for the safety of journalists.282 This is in part in response to the incidence of threats of violence and rape against journalists.283
3.100 It was our provisional view that the general harm-based communications provided sufficient coverage of “threatening” and “menacing” communications.284 However, given the potentially heightened harm that genuine threats pose, we asked a specific consultation question on threatening communications:285
In addition to our proposed new communications offence, should there be a specific offence covering threatening communications?
3.101 Professor Alisdair Gillespie argued that while threatening messages may fall within the scope of the general communications offence, threatening behaviour is distinct behaviour with which the criminal law should concern itself, and that a separate offence would be appropriate:286
On the one hand, threatening messages would seem ordinarily to come within the proposed offence. It is easy to imagine how someone threatened could suffer serious emotional distress. However, they won’t automatically suffer such harm. Indeed, they may not personally feel threatened even though it is undoubtedly a threatening email.
Making threats is something the law should concern itself with. Threats in person can constitute a (technical) assault, although it would be difficult for them to do so online (as the person would not apprehend immediate unlawful violence). Public order law also criminalises threatening behaviour in various circumstances.
It would be perfectly appropriate to define an offence of sending a threatening communication. Chambers may be cited as a reason why not to, but it isn’t such a reason. That was an example of bad cases make bad law. The case should never have been prosecuted in the first place, as it was evident from the outset that it was not a true threat. Its importance should not be overstated in this arena.
3.102 The Association of Police and Crime Commissioners agreed, noting that while the behaviour may be covered by a general offence, this may demonstrate a labelling problem given the seriousness of the conduct:287
Whilst we recognise that existing communications offences explicitly cover threats, and that threatening and menacing communications will also be covered by this proposed offence, we do believe that in cases where a threat is only covered by the proposed new communications offence, that there may be a labelling problem.
As commissioners of victims services, we recognise the impact that a communication that contains threatening or menacing language may have on victims, and would argue that in order for the offence to adequately capture and describe the nature of wrongdoing, there is sufficient justification for a specific offence relating to threats to be created.
3.103 Refuge argued for greater clarity in the use of existing offences, and distinguished culpable threatening behaviour from that more appropriately addressed by a general communications offence:288
Refuge.. .supports specific threats being criminalised where there is clear need. For example, we are campaigning for threatening to share intimate images without consent to be criminalised through the extension of section 33 of Criminal Justice and Courts Act 2015 to include threats as well as the sharing of an intimate image or film. This issue falls within the scope of a separate Law Commission project and therefore we do not go into detail here. However, the full details of our proposal can be found in our report The Naked Threat.
Many existing offences cover various threats, for example threatening to kill, harassment offences and the coercive and controlling behaviour offence. In our experience these are under utilised by the police and CPS. Whilst prosecutions for tech abuse related offences are relatively rare, where the police do take action, they usually favour malicious communications offences, even when the communication was a threat to kill or a threat as part of a pattern or stalking and harassment. It would be our preference for the existing offences to be used more widely, as they better reflect the context and seriousness of a threat compared to a communications offence and therefore attract a more severe sentence.
3.104 English PEN agreed that a specific offence addressing threatening communications would be appropriate, but cautioned that any offence must address genuine threats:289
So long as measures are proportionate, we believe that preventing threats to individuals is an appropriate reason to curb Article 10 rights — threats of violence represent a significant interference with an individual’s Article 5 right to security and the Article 8 right to a private and family life.
Such an offence would need to be drafted to take into account the infamous ‘Twitter joke trial’ (Chambers v Director of Public Prosecutions [2013] 1 All ER 149), which the Court of Appeal acknowledged should not have been prosecuted. This could be addressed with a strong mens rea element that rules out jokes and unfocused rants that happen to include threatening language.
Once more we note how a message ‘broadcast’ over social media to no-one in particular is conceptually different to a message addressed to a particular recipient. This is what distinguishes the Chambers case from the prosecution of Matthew Wain (discussed at paragraphs 5.207-5.208).
3.105 Suzy Lamplugh Trust described the existing offences that address threats, noting that a broader offence could ensure that the varying types that are not covered do not go un-criminalised:290
Although some threats are covered in other acts (such as the threat to kill under section 16 of the Offences Against the Person Act 1861) there are several threats that may not be covered. Namely those that are likely to cause harm within the context of abuse at work, by members of the public or colleagues (threats to share intimate images, personal information, etc); and likewise within the context of stalking where the behaviours are so varied and not listed as offences within themselves (threat to visit victim’s house, threat to contact their family member, etc).
3.106 Fix the Glitch agreed, noting the heightened harm that threatening messages pose:291
We believe the proposal should include a specific offence covering threatening communications. As a charity working to end online abuse, we recognise that online violence takes on several forms, ranging from harmful to illegal and threatening content. Threatening content - including through images and video - poses a direct danger to a person’s integrity and/or life, or their ability to exercise their rights freely. Research has shown, for instance, that politicians who have faced threats online in the UK have been forced to step down from some of their activities. While abusive content has a long-term psychological and emotional impact on victims - and can lead to threats further down the line - threats pose an immediate danger to the victim and need to be addressed separately in legislation.
3.107 Dr Zhraa Alhaboby agreed, noting that her work with victims of online abuse often involved discussing specifically threatening messages: 292
Threatening communication was one of the contributing factors to reporting to the police when we examined the patterns of reporting in police records. We examined disability-related cyber incidents in UK police force records over 18 months, and identified three overarching themes as influencers of reporting or documentation of cyber incidents by the police, these were: psychological impact, fear for safety, and the type of disability.
In my research threatening communication was commonly reported with fear, which is significantly associated with cyber-victimisation.
3.108 Mr Justice Martin Spencer, in his capacity as chair of the Judicial Security Committee, argued that a separate offence is necessary given the incidence of threatening judges:293
I think it should be wider and include threats because of the issues with judges. Threats are a problem. It would be easier if it were more specific.
3.109 The LGB alliance noted that while they were opposed to the other proposals, a specific offence to address threats would be welcome:294
This is the only addition that we feel would be a helpful step forwards. People in high-profile positions, especially parliamentarians and academics and especially women, receive rape and death threats on a daily basis. Although these are already covered by existing legislation, we do think it is worthwhile considering adding a specific offence covering such threats.
3.110 Refuge explained some of their work with victims of threats, noting the difficult in some cases of ensuring threats are specifically addressed with more serious offences. They noted that their preference would be to see better outcomes via the existing offences addressing threatening behaviour rather than a specific offence:295
For example, we recently supported a survivor whose perpetrator made repeated threats to kill her. She reported this to the police and on several occasions and no further action was taken. The final time she reported the threats to the police, a case of malicious communication was pursued, and he was charged. The perpetrator went on to attack the survivor and left her with life changing injuries.
In another case a perpetrator shared a sexual image of a survivor without her consent. The survivor report this to the police and the perpetrator was charged under the Malicious Communications Act. Refuge staff questioned why a charge under section 33 of the Criminal Justice and Courts Act was not pursued instead, and received the response that the police try to avoid using the more serious offences as these charges are often rejected by the CPS.
Based on our experience we think addressing gaps in the law regarding specific threats and focusing on increasing the use of existing offences covering threats would lead to better outcomes for survivors than a general threatening communications offence.
3.111 Women’s Aid stressed the importance of addressing threats:296
We would argue that it is important for there to be a specific offence covering threatening communications. Threatening communications should not be downplayed or dismissed, particular when raised by women, survivors of VAWG or other marginalised groups. The response to threatening communications must reflect the reality of the impact that they have and the seriousness of the intention behind them.
3.112 The Antisemitism Policy Trust referenced the prevalence of threats to parliamentarians and their staff in supporting the proposal:297
Yes. The Trust’s experience of working to support parliamentarians and their staff, in particular, leads us to believe threatening communications have a particular, significant harm which should be separated from a general offence. Threatening communications can and do impact people’s way of life, their fundamental freedoms.
3.113 Suzy Lamplugh Trust expressed their support and noted the increased incidence of threats during the COVID-19 pandemic, as well as the potentially narrow scope of the section 16 OAPA 1861 offence:298
Suzy Lamplugh Trust welcomes the proposal for a specific offence covering threatening communications. Inflicting a comparable degree of psychological and emotional harm as those offences at the centre of this consultation, we argue that it is equally important to try and convict those sending or posting threatening messages online. Especially within the context of the COVID-19 pandemic and the consequent increase in online communication, threats among frontline workers such as healthcare and retail workers have risen significantly.
3.114 ARTICLE 19 agreed that, to the extent it is not already criminalised, an offence addressing credible threats of bodily harm could be worth investigating.299
3.115 The Criminal Bar Association disagreed, arguing the general communications offence would be sufficient:300
We consider that [the general harm-based communications] offence would be sufficient to obviate the need for a further offence. Should there be significant feedback to suggest that such an offence is desirable, it would be possible to add a further subsection to provide for an aggravated form of the offence.
3.116 Fair Cop answered “No”:301
We do not see a need for this to be a separate offence. If a malicious communication is directly threatening to an individual then this makes it more likely that it has genuinely caused significant distress and less likely that the person publishing it could have any good reason. We consider that communications that directly threaten an identifiable individual are very likely to justify the intervention of the criminal law.
3.117 The Free Speech Union disagreed, indicating their preference to limit the number of speech offences:302
No. The Commission itself points out at Para.5.210 that there seems no particular difficulty arising from the lack of any offence of this kind. Given this fact, we for our part see no need whatever to add to the catalogue of speech offences.
3.118 The Crown Prosecution Service noted that external communications about prosecutions for general offences could address the labelling issue to an extent, and that other specific offences do exist that could be used to address threats:303
As a matter of law, threatening communications may be sufficiently covered by the proposed harmful communications offence.
We note the Law Commission welcomes our views on whether the labelling issue is a sufficient justification for a specific offence relating to threats. This is something that the CPS would seek to assist in addressing through effective external communications following successful prosecution outcomes.
Existing offences also cover threats e.g. [section]16 of the Offences Against the Person Act 1861 [threats to kill] and [section] 51 of the Criminal Law Act 1977 [bomb threats].
3.119 There was extremely strong support from consultees for a separate offence that specifically covers threatening communications. This support came both from those who were in favour of the provisionally proposed harm-based offences and from some who opposed them. In our view, this reflects the fact that genuinely threatening communications reflect a high degree of criminal culpability.
3.120 Some consultees noted that there may be a gap created via the harm-based offences wherein a communication is sent or posted that may be properly characterised as threatening but would not meet the threshold of harm. None provided examples, and it seems unlikely that a genuine threat would not meet the threshold of harm in the provisionally proposed offence.
3.121 A recurring comment in consultee responses was that deliberate threats constitute a particular harm that would not be sufficiently addressed via a generic communications offence. Instead, in their view, they should be addressed by a fairly labelled offence. Similarly, the maximum penalty of two years imprisonment would apply for a credible threat of an offence other than murder (for example, of rape or of some other serious violent offence). 304Consultees also raised the potential links between online abuse and offline violence in the context of threatening communications as a reason to address this matter explicitly.
3.122 While the matter of Chambers305 was raised by some consultees, none thought it posed an insurmountable problem, and noted that providing the offence targeted genuine threats, it should not stand in the way of addressing the harm. We agree with Professor Gillespie’s characterisation of it as involving no true threat.
3.123 Refuge noted the poor application of existing offences that target threats. They disagreed with the need for a specific offence to tackle threatening communications and instead urged for greater focus on the communications offences more broadly. A similar point was raised by the Suzy Lamplugh Trust, who argued that certain threatening behaviours manifest in otherwise innocuous behaviours. We appreciate the complexity of addressing these threatening behaviours, and that any specific offence may not explicitly address these concerns. It is worth re-iterating the availability of harassment offences for courses of conduct (that may otherwise seem innocuous).306 Further, as discussed in Part 2 above in relation to persistent communications, the various specific protections contained in the Domestic Abuse Bill will be vital to ensuring there is a comprehensive strategy to address violence against women and girls.307
3.124 The consultees who disagreed tended to note that the current lack of a specific offence was not problematic, and that in any event our provisionally proposed offence (and to a lesser extent the existing communications offences) cover this conduct. They did not offer specific responses to the fair labelling problem. Some suggested an aggravated form of the general offence. Again, we do not believe this adequately addresses the labelling problem, nor reflects the distinct harm caused by certain types of serious threatening communications.
3.125 Threats, and specifically threats in sexual offences, were considered in detail in our consultation paper on Intimate Image Abuse.308 Threats in the context of intimate image abuse are a particularly complex matter, which have been dealt with in various ways in other jurisdictions. 309As we set out in the Intimate Image Abuse consultation paper, we provisionally concluded that a specific offence criminalising threats in the context of intimate image abuse was warranted.310 While there are a range of existing offences that could criminalise certain types of intimate image threats, it was our view that they were not always sufficient or appropriate. We ultimately made a provisional proposal to criminalise threatening to share intimate images by way of a specific offence.311
3.126 While the threats subject to our provisional proposals in the Intimate Image Abuse project distinctly relate to sexual offending, and would correspondingly be subject to a range of ancillary orders and procedural protections, 312our concern here is with threats of serious harm more generally. The harm at issue is that which attaches to threats as a distinct form of communication, as consultees set out in their responses and we discussed at paragraphs 3.101 to 3.105 above.
3.127 After considering the detailed responses consultees provided, in our view it is appropriate to have a specific offence addressing genuinely threatening communications, in particular targeting threats to cause serious harm. In determining how best to define the level of harm, we have considered both our analysis of threats in relation to Offences Against the Person, 313and the discussion about consistency in setting equivalent thresholds of harm in our report on Misconduct in Public Office.314
We believe that the threats within scope should be threats to cause serious harm, where “serious harm” includes threats of causing serious injury (encompassing physical and psychiatric injury within the definition of “grievous bodily harm” for the purposes of sections 18 and 20 of the Offences Against the Person Act 1861315), rape or serious financial harm.
3.128 In determining the differing types of harm within the scope of the offence, we were guided by the responses we received, some of which are extracted above. Communications that threaten serious harm relate to physical and psychiatric injury, and threats to rape relate to violence and sexual violence respectively. The seriousness with which the law treats this behaviour is reflected in the substantive offences contained in the Offences Against the Person Act 1861 and Sexual Offences Act 2003. We accept that serious financial harm is of a different nature to the first two categories of harm. However, it is also an area that the criminal law addresses substantively; as we reference in Part 1 above and in Chapter 2 we recognise that certain types of financial harm are the subject of specific criminal offences (such as those in the Fraud Act 2006).
3.129 The harm posed by threats of financial harm (including, as we noted in the consultation paper, “financial ruin”316) can be distinguished from the “knock-on” financial impact that may result from generally harmful communications within scope of the offence in Chapter 2 or the knowingly false harmful communications offence in Part 1 above. We are conscious that there are existing offences that deal with the substance of financial harm (including section 2 of the Fraud Act 2006) or threatening behaviour in certain circumstances (for example the offence of blackmail in section 21 of the Theft Act 1968). However, just as the substantive offences in the Offences Against the Person Act 1861 and Sexual Offences Act 2006 stand alone, communications that convey threats that relate to the substantive behaviour covered in those Acts are in our view a matter that, in the words of Professor Gillespie, “the law should concern itself with.”317
3.130 We believe that defining “serious harm” in this way would set a seriousness threshold in respect of each type of harm that is consistent with other areas of the criminal law, and therefore readily applicable in practice. The seriousness threshold would also ensure that only sufficiently culpable behaviour is within the scope of the offence. Further, the offence would complement the other offences we recommend elsewhere in this report, forming a suite of communications offences that can be applied to distinct types of harmful communications.
3.131 We are also of the view that given the seriousness of the harm, it is appropriate to recommend that the fault element of the offence captures both intention and recklessness, similar to our provisionally proposed offence of threatening to share an intimate image of another.318 We acknowledge this may appear inconsistent with the requirement for intention in the harm-based offence in Chapter 2: . However, as we set out above, threatening communications are a particularly harmful subset of communications. This offence is restricted to threatening communications, meaning its scope is far narrower than the harm-based offence, which applies generally. Moreover, it the threshold of “serious harm” means that it will only apply to communications that contain genuine threats of grievous bodily harm, serious financial harm or rape. For these reasons we do not believe that there is any meaningful inconsistency between the harm-based offence and this approach to a distinct offence targeting threatening communications.
3.132 Another matter to consider is whether the offence should contain any explicit statutory defences, given similar existing provisions. 319We do not believe that a person could ever claim to reasonably believe that a threatening communication conveying a threat of grievous bodily harm or rape could be justified as a proper means of justifying a reasonable demand. However, we are conscious that serious financial harm is an appreciably different type of harm. We believe it is appropriate to include a defence where a threatening communication conveying a threat of serious financial harm is made by a person to reinforce a reasonable demand and they reasonably believed that the use of the threat was a proper means of reinforcing the demand. However, in the event a defendant sends a threatening communication which conveys a threat of serious harm amounting to grievous bodily harm or rape in circumstances of duress or necessity, the defences of duress or necessity would be available in the ordinary way.320
3.133 While we do not recommend a specific penalty, we believe that any offence specifically addressing threatening communications would appropriately have a penalty that sits between the “either-way” harm-based communications offence that we recommend (with a maximum penalty of two years’ imprisonment) and the maximum penalty of 10 years’ imprisonment for the threats to kill offence in section 16 of the Offences Against the Person Act 1861. 321The resulting offences would exist in a complementary hierarchy allowing increasingly criminally culpable behaviour to be specifically sanctioned. This would reflect the increased culpability of specific, genuinely threatening communications as compared to generally harmful communications (which may include less serious threats that fail to satisfy the threshold of the specific offence).
3.134 In line with our recommendations made on offences against the person, 322and concordant with the extremely strong consultee support for a specific offence addressing threatening communications, we are of the view that threatening communications should be explicitly addressed with a specific either-way offence.
3.135 We recommend that a separate either-way offence explicitly addressing genuinely threatening communications be enacted with the following elements:
(1) the defendant sent or posted a communication which conveys a threat of serious harm;
(2) in conveying the threat:
(a) the defendant intended the object of the threat to fear that the threat would be carried out; or
(b) the defendant was reckless as to whether the object of the threat would fear that the threat would be carried out.
(3) The defendant would have a defence if he or she could show that a threat to cause serious financial harm was used to reinforce a reasonable demand and he or she reasonably believed that the use of the threat was a proper means of reinforcing the demand.
(4) For the purpose of this offence:
(a) a communication is a letter, electronic communication or article (of any description); and
(b) serious harm includes serious injury (amounting to grievous bodily harm as understood under the Offences Against the Person Act 1861), rape and serious financial harm.
3.136 One particularly troubling behaviour that was reported to us over the course of this project was the intentional sending of flashing images to individuals known to have photosensitive epilepsy. The Epilepsy Society reported that, on 13 May 2020, their Twitter account, with its 26,000 followers, was attacked with over 200 posts over the course of 48 hours. The posts including flashing GIFs and violent and pornographic content. Many individual followers were personally targeted.323
3.137 This behaviour will almost certainly fall within the scope of our harm-based offence. The Epilepsy Society, in their consultation response, confirmed that psychological harm is one aspect of the harm suffered by those who are targeted with flashing images. 324They outlined the serious distress that people with photosensitive epilepsy who had been targeted with this behaviour and their families experienced, ranging from the triggering of seizures after prolonged periods of being free of seizures to the serious distress experienced when reading the abusive messages that often accompanied the flashing GIFs and images. Epilepsy Society also note that those with epilepsy - but not photosensitive epilepsy - are also targeted, and that this causes “very real psychological harm”. However, when a seizure is caused, the potential seriousness of the offence would seem to warrant a more serious response other than a general communications offence.
3.138 It was our view in the consultation paper that such behaviour ought ideally to be prosecuted as an offence under section 18 of the Offences Against the Person Act 1861 (maliciously causing grievous bodily harm or wounding, with intent to cause grievous bodily harm), or section 20 of the same Act (maliciously inflicting grievous bodily harm or wounding). 325However, we do acknowledge that this involves two prosecutorial hurdles in particular. First, it might be difficult in all cases to establish that the harm met the requisite threshold of “really serious harm”. A seizure, without more, does not automatically meet that threshold (though clearly a seizure brings with it substantial risk of a multitude of serious harms). 326Secondly, establishing causation beyond reasonable doubt might prove difficult, especially if the person is known to suffer from regular seizures. Clearly it will not always be difficult - in some cases it may be very obvious that the flashing image caused the seizure - but we cannot guarantee that there could never be reasonable doubt as to causation.
3.139 The further difficulty with reliance on the Offences Against the Person Act 1861 is that, as a result of the wording of the offences, the only offences that could be charged are the most serious, namely sections 18 and 20, because these require only that grievous bodily harm or wounding be caused or inflicted. The lesser offence, under section 47, is assault occasioning actual bodily harm: as the name suggests, the offence requires proof of an assault or battery. Despite the lower threshold of harm, proving that the flashing image (without more) amounted to an assault or battery will be very difficult.
3.140 Notably, this shortcoming of section 47 is addressed in our earlier recommendations for reform of offences against the person. We recommended replacing section 47 with an offence of intentionally or recklessly causing injury (whether or not by assault).327 The physical harm threshold is sufficiently low in that offence that seizures would certainly satisfy that element of the offence. This does not, of course, address the difficulties of proving causation.
3.141 However, in the absence of such an offence in statute, and recognising the difficulties of proving causation in some cases, we therefore consider that there is real scope for an offence specifically targeted at the intentional sending of flashing images to those with epilepsy. Having not proposed such an offence in the consultation paper, and
therefore having not consulted on its terms, we do not recommend a detailed offence.
We do, however, recommend that an offence should be created.
3.142 We recommend that the intentional sending of flashing images to a person with epilepsy with the intention to cause that person to have a seizure should be made an offence.
4.1 We set out in the consultation paper that we did not intend for the proposed “harmbased” offence to cover regulated media such as news media, broadcast media and cinema.1 While we expect that the elements of the offence itself would ensure that all communications made by the media in good faith would not be captured, in that they would not be sent or posted with an intention of harming a likely audience, and that they would be being sent or posted with a “reasonable excuse”, we accept that an explicit exemption will mitigate any potential residual “chilling effect”.
4.2 There are various justifications for explicit exclusion of the media. For example, not to do so may constitute a dramatic expansion of the scope of the existing communications offences, which is not consistent with our Terms of Reference. It is further worth bearing in mind the crucial nature of a free and independent press in a democratic society, and the various existing regulatory schemes within the media.2
4.3 For the avoidance of doubt, we did not propose extending the exemption to comments below articles (though, as we note below, there are safeguards in the offence itself to protect communications on matters of public interest).
4.4 As we set out below, although we initially only proposed an exemption for the general harm-based communications offence, consultees urged us to consider it in the context of both that and the false communications offence.
4.5 We asked a specific consultation question on the press exemption.3
4.6 Consultees generally expressed strong support for an exemption for news media, broadcast media and cinema. However, many did raise concerns or questions about how the exception would operate.
4.7 Hacked Off disagreed with the need for an exemption. They set out their view that the news media should be covered by any new offence. In responding they raised concerns with the nature of existing regulatory regimes for the media and discussed the nature of protection for freedom of expression in society beyond the press. They argued that any “chilling” effect of a new offence on the news media should not be any less than on society more broadly. Their response expressed their desire to ensure
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 5.66.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 5.67. As we noted at footnote 386 of the consultation paper: “Newspapers and magazines are primarily regulated by IMPRESS and the Independent Press Standards Organisation. Television and radio are regulated by Ofcom. Cinema is regulated by the British Board of Film Classification.”
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 2.
that any new laws should apply without differentiation. We extract a number of their key points:328
It is our view that news media should be covered by the new offence. We believe that the arguments given by the Commission in favour of exempting news media from the offence are flawed. Further, there are good reasons for the news media to be covered.
The first argument given in favour of exempting the news media is that it is “regulated” (paragraph 5.66). This is not correct in relation to the large majority of print & online news media. Of national titles, some rely on a “readers’ editor” model and others rely on the IPSO.
The second argument given is that “a free and independent press is crucial in a democratic society” (paragraph 5.67). It is undoubtedly true that freedom of expression as exercised by some media outlets is crucial in a democratic society. But, as is stated in Article 10(2) of the European Convention on Human Rights, the exercise of this freedom carries with it duties and responsibilities including those which protect the rights of others. The news media’s right to freedom of expression is subject to those duties and responsibilities in the same way as the freedom of expression of any citizen...
The third argument is that “there is a risk that mere existence of the offence would have a chilling effect on the exercise of freedom of expression by the press and media outlets” (paragraph 5.67) .. A “chilling effect” will, of course, be felt equally by ordinary citizens, a result which is implicitly accepted as an appropriate one. There is no reason why the news media should be in any different position. Unregulated news media outlets should not be treated as if they have privileged access to the right of free expression. There is no basis for such special treatment.
4.8 Gina Miller expressed her concern with a media exemption. She noted that her experience of online abuse often related to comment sections under news articles. She also discussed the slightly separate issue of private and encrypted messaging:329
I question why the news and broadcast media are not covered.
Much of the abuse, harassment and hate targeting me / I received occurred in comment sections under online news articles.
Several complaints made on my [behalf] to IPSO regarding comments made by the public that contained threats of violence and abuse failed as the IPSO code stipulates that such comments are outside their regulatory scope. This is despite the fact that IPSO acknowledged these comments caused me distress.
I would argue that now that newspapers are online, and the Government is intending that OFCOM is the regulator for online harm, that newspapers and broadcast media have to moderate and exercise a duty of care on comments posted by their readers/audience. The issue should not be the medium but the messages.
4.9 The Media Lawyers’ Association provided a detailed submission in support of an exemption:330
We agree that news and broadcasting media should be expressly exempted from the proposed new offence. We concur with the Law Commission’s explanation at paragraph 5.67 of why this is so important.
It is correct to state, and we emphasise, that “the mere existence of the offence would have a chilling effect on the exercise of freedom of expression by the press and media outlets”. And it is correct to note that, “the existing regulator[y] framework within which news media and broadcast media groups operate render the imposition of additional criminal liability superfluous”. There is no need for further regulation of an already regulated industry.
In short, a clear and broad exemption for journalistic material is required and the Law Commission identifies considerations which are critical in determining the proper scope of such an exemption.
4.10 Similar arguments in favour of an exemption were outlined in responses we received from the News Media Association and DMG Media.331
4.11 The Criminal Bar Association agreed with our proposal, suggesting that any exception should be included in the Act itself, and that it should cover journalistic activity beyond purely published material:332
Reference is made in the consultation to the exclusion of news media etc being achieved ‘by way of a carve out if necessary’. We would suggest that it is necessary for a term excluding these forms of communications from being covered by the offence to be in the Act itself.
We would also suggest that the exclusion should be broad enough to cover communications that are not necessarily intended for broadcast in the form in which they have been sent: that is, a journalist who sends to a news corporation via email some working document/reports or a series of videos designed to provide an editor with background material, but not intended for broadcast in the form sent, should not be caught by this offence. We would suggest that an explicit statement of the type of materials/situations the new offence is not intended to cover should be made, rather than causing journalists to have to consider if they have a reasonable excuse in relation to briefing materials in each individual case.
4.12 The Crown Prosecution Service observed that an exemption would reflect the position under the existing communications offences:333
There is merit to the Law Commission’s proposal to exclude news media, broadcast media and cinema from the proposed harmful communications offence. We understand that the offence would deal with individuals and not with regulated news media. Broadcast media were also excluded from s.127 (4) CA 2003, and criminal libel was abolished in 2010.
4.13 IMPRESS expressed their support for a media exemption, but urged caution with respect to exempting sections of the media that do not submit themselves to public interest regulation. They set out their view that an exemption “should be designed to both affirm press freedom while at the same time [protect] the public from harm.”334
4.14 The Association of Police and Crime Commissioners agreed with the exemption of the media, noting the existing regulatory schemes:335
We agree that these offences should not cover regulated media such as the news media, broadcast media or cinema, and that, as the Commission state, this could potentially have a chilling effect on the exercise of freedom of expression by the press and media outlets. We also note that regulatory schemes already exist for these media.
4.15 Fix the Glitch agreed that the focus should not be on traditional media, and emphasised the growth of online abuse on social media:336
We agree that this offence.. .should exclude broadcast media, news media or cinema. This allows the law to consider electronic communications in their specificity and the impact of harmful communications made on social media, the Internet and online platforms, while keeping up with the technological advances and changes made to platforms on a regular basis. Harmful content on the Internet and social media is characterised by its growing scale and the ease with which it is disseminated and can be shared. Since its launch in 2017, Glitch has documented the scale and nature of online abuse targeting women and minoritised communities. We have found that a significant proportion of women have faced online abuse. During the COVID-19 pandemic, our research showed that one in two women and non-binary social media user[s] experienced online abuse. Given the sheer scale of online abuse and its many manifestations, current legislation is not adapted to respond to the consequences of online harms.
4.16 The National Secular Society supported an exception for the media but noted the tension that then arises when individuals seek to present their own views:337
We agree that freedom of expression should be protected for news media, broadcast media and cinema. However, we are concerned that this approach gives freedom of expression to those who can present their views in newspapers or on television whilst denying it to members of the public who lack such a platform and instead use YouTube or a personal blog. Likewise, it would protect film studios but not artists who create low-budget productions and share them via social media. This is an elitist approach and could lead to a situation where a news outlet can lawfully publish an article but a person who shares it on social media and echoes the sentiments expressed therein could face prosecution.
4.17 The Free Speech Union agreed with the arguments in favour of a media exemption but noted that media regulation is itself problematic, and that an approach that only referred to “regulated media” may not be sufficient:338
We wholeheartedly agree that the news media, broadcast media and cinema need to be protected here, and would support a carve-out.
We have, however, one comment. We note that the Consultation Paper refers to “regulated” media (para.5.66). If this means that the media carve-out should be limited to those media that are in fact regulated, we regard this as problematic. In particular, it has to be remembered that not all newspapers are regulated (membership of an organisation such as IPSO or IMPRESS is in law voluntary), and IP broadcasts originating from outside Europe are entirely unregulated (since Ofcom’s writ does not run there). Since we regard it as important that free speech protection extend to unregulated as much as regulated speech, we would press for any protection to apply whether or not the relevant publication is otherwise regulated in any other way.
4.18 Mermaids welcomed the proposal to exempt media. They also supported our position in the consultation paper regarding comments on news articles:339
We agree and echo the Law Commission’s hesitancy to ‘dramatically expand the scope of the existing communications offences’, and strongly agree that a ‘free and independent press is crucial in a democratic society’, and that the ‘burden of justification for imposing criminal liability...is especially weighty’ (para. 5.66-5.67 of consultation paper).
We support the Law Commission’s proposal that the new offence would cover any abusive comment posted in response to an article. We see first-hand at Mermaids that often replies and comments to posts often contain the very harmful communication.
4.19 The Crown Prosecution Service noted the potential problems that exist in defining “news media” and “broadcast media” further:340
We consider that the 'news media' and 'broadcast media' should be further defined with further consideration of newer forms of news media online such as blogs, user generated content, and comments on articles. Although ‘broadcast media’ is defined in some statute, it may be less straightforward to define ‘news media’. To that end, further consideration of the definition of 'news media' and 'broadcast media' would be welcomed.
4.20 The Bar Council agreed with the proposal subject to a question about how “media” is defined:341
We agree, subject to identification of what is meant by the “media”. Would the exclusion include a communication in the form of a newsletter from a malicious group intent on disseminating false, alarming and foreseeably harmful information?
4.21 The News Media Association provided a detailed response, which set out various possible definitions of an exemption. They emphasised that any exemption must be “clear, comprehensive and robust”. The response also set out their view that any false harmful communications offence should also have a exemption in the same terms.342
4.22 The Media Lawyers’ Association noted the exemption the Government outlined in its response to the Online Harms white paper, arguing that our proposal in the consultation paper was not sufficiently broad:343
...it appears that the Law Commission is proposing that its exemption will be limited to articles published on news sites. This is far too narrow and fails to provide the necessary protection for freedom of expression which the Law Commission aims to achieve through its reforms. It is also inconsistent with the approach taken by the Government in its recent Online Harms White Paper: Full Government Response to the Consultation.
An exemption must include all content on news sites and must extend to that content when it is shared and when it appears on third party sites including social media platforms. It must also include user comments and social media. It must extend to communications sent for journalistic purposes, including, for example, communications offering a right to reply.
This approach has been accepted by the Government in its recent Online Harms White Paper: Full Government Response (“Online Harms: Full Response”). In the Executive Summary, the Government states (our emphasis added):
Journalistic content
22. ... Content published by a news publisher on its own site (e.g. on a newspaper or broadcaster’s website) will not be in scope of the regulatory framework and user comments on that content will be exempted.
23. In order to protect media freedom, legislation will include robust protections for journalistic content shared on in-scope services. The government is committed to defending the invaluable role of a free media and is clear that online safety measures must do this. The government will continue to engage with a range of stakeholders to develop our proposals.
And in Part 1 the Government states its Final Policy Position as follow[s]:
“Content and articles produced and published by news websites on their own sites, and below-the-line comments published on these sites, will not be in scope of legislation. In order to protect media freedom, legislation will include robust protections for journalistic content shared on in-scope services. The government is committed to defending the invaluable role of a free media and is clear that online safety measures must do this. The government will continue to engage with a range of stakeholders to develop these proposals.”
We agree with this approach and submit that the exemption from any reform communications offence(s) must be equally broad.
They went on to describe existing provisions that exempt journalism and journalistic activity:344
^we simply note that there are templates in existing legislation including, by way of example, exemptions and defences referring to “the purposes of journalism” and “journalistic material” contained in the Data Protection Act 2018, the Police and Criminal Evidence Act 1984 and the Investigatory Powers Act 2016.
4.23 Community Security Trust noted the potential problems with simply adopting existing media regulatory regimes as a way of exempting the media:345
Though IMPRESS and IPSO as regulatory organisations are important in regulating its members, there are many online 'news media' outlets that may have print operations who are not regulated.
Examples include:
- Heritage and Destiny (http://www.heritageanddestiny.com/) - The online manifestation of a UK print publication that, in its own words 'reflects a cross-section of 21st century racial nationalist opinion'.
- A news website example is Dorset Eye (https://dorseteye.com/) - An 'independent not for profit news website' which includes 'Dorset News' but has been alleged to publish articles with antisemitic content and that target Jewish people with hate (https://antisemitism.org/dorset-eye-community-website-used-as-cover-to-spread-
antisemitic-hate/ and
https://twitter.com/sffakenews/status/1157994330646339584?lang=en).346
Any legislation also needs to cover news media that has an online output, not covered by existing regulators.
4.24 Carnegie UK Trust agreed that the media deserved differential treatment. However, they noted that technological development has made a clear definition of “media” more difficult. They ultimately argue that rather than an explicit exemption, the media should be best dealt with through the public interest element of the offence:347
We agree that the media deserves special consideration because of its role in informing the public and holding those in power to account. This process may give rise to considerable discomfort or distress to those the subject of reporting, but it would rarely be in the public interest to criminalise this. The report notes in passing the possibility of a ‘carve out’. In this context we note that, given the development of online distribution channels, it is difficult to define the different elements of the media clearly. Referring to regulated media would not suffice. While there are regulatory regimes, not all the traditional media are members of such a regime (see e.g. The Guardian, the Financial Times). Choices made by the print media to remain selfregulated and outside a Leveson framework means that there is not a simple definition. Moreover, the democratisation of public speech means that actors beyond the traditional media actors may speak to large audiences in the public interest. This can be seen most clearly in prominent social media accounts that have more followers than major newspapers have readers. Rather than seek to define an exception relating to the media, which would quickly run into definitional difficulties and potentially exclude other forms of public interest speech, our view is that the position of public interest speech, as usually found in the media, could be best dealt with through a public interest element to the offence, as proposed in 5.51(6).
4.25 We are grateful for the detailed responses outlined above. This was an understandably contested aspect of the consultation. Nonetheless, we remain of the view that some kind of exemption is required. The existing communications offences were not intended to regulate the press. That the offence in section 127 of the Communications Act 2003 now does apply to the press is by accident of technological development rather than by design: it was not envisaged in 2003 that, less than a decade later, the vast majority of news and media consumption would be via a “public electronic communications network”. As a result of how journalism and news media is consumed, the offence that we propose would, absent an exemption, apply to most, if not all journalistic output (every single article and video posted online). It would be a form of criminal law regulation of press content.
4.26 Of course, we are not arguing that journalists should be free to commit criminal offences under the guise of journalistic freedom. Our position is that the readership of much of the news media is so broad, and their content so extensive, that some harm resulting from content becomes very likely (perhaps so much so that it be deemed intended348). A lot of reliance therefore has to be placed on the “lack of reasonable excuse” element; was the article “reasonable” given its journalistic purpose? This may have a chilling effect on freedom of the press, and is undoubtedly a form of criminal press regulation by another name. We have been asked to review the criminal law provisions for abusive communications; we do not think it appropriate to impose press regulation by a side-wind.
4.27 It is worth noting here, as we raised in the Introduction and expanded on at length in Chapter 2, that the shift away from requiring only an awareness of a risk toward requiring intention means that some of the concerns raised by consultees in relation to our harm-based offence are lessened. However, an explicit exemption will mean that notwithstanding the more constrained scope, it will be beyond doubt that the offence will not apply to press content.
4.28 We are also sympathetic to the view that, to avoid any “chilling effect” that our recommendations regarding knowingly false communications could have, that the exemption applies to both offences. Again, while we are confident that the elements of that offence are such that it would not apply to regulated media such as the news media, broadcast media, and cinema, we are sympathetic to consultees’ views that the absence of a clear, comprehensive exemption could have an undue “chilling effect” on the freedom of the press.
4.29 We also believe that an explicit exemption for both the general harm-based communications offence and the knowingly false communications offence is the most coherent with the Government’s approach to the regulation of online harms.349
4.30 In relation to comments made on articles published online, we were not persuaded by those consultees who thought any exemption should also extend to comments. Conversely, our proposal not to exempt comments was met with strong support, and a number of consultees outlined their own experiences of abuse that arose from comments on news articles. We note that excluding comments from the exemption does not deprive them of the safeguards inherent in the offences themselves. These include the requirement that the prosecution prove beyond reasonable doubt that a communication was sent or posted with an intention to cause harm to a likely audience and that it was sent without reasonable excuse.
4.31 As consultees set out, the lack of a comprehensive regulatory regime that covers all news media, broadcast media, and cinema means that any exemption must look beyond any single type of media regulation. For that reason, in our view an exemption along the lines of the wording adopted in the Draft Online Safety Bill 350relating to protections for journalistic content should be replicated in relation both to the harmbased offence and the harmful false communications offence.351 We believe that this approach will ensure press freedom, above and beyond the protections afforded by the elements of the offences (including the fault element and the “reasonable excuse” safeguard), while simultaneously protecting against harmful communications.
4.32 At the time of writing, clauses 14(8), 39 and 40 of the Draft Online Safety Bill provide definitions of “journalistic content”, news publisher content and recognised newspublisher and news-related material. We accept the provisions are only in draft form. However, we see benefit in consistency in this area. In our view, the draft wording appears to reflect the context and purpose of the exemption we believe appropriate for the harm-based and knowingly false communications offences, save for the issue of comments made on news articles online. It also appears to avoid the difficulties with our provisionally proposed approach that only exempted regulated media which were raised with us by consultees.
4.33 Expanding on the rationale for including comments made on news articles online in the scope of the offences: we note that there is a divergence in purpose between the regulatory regime set out in the Draft Online Safety Bill and our recommended offences. The latter exist to sanction criminal conduct rather than censor or promote industry regulation. In our view, this divergence in purpose is significant in the context of comments made on news articles. It is understandable that a regulatory regime may not seek to make publishers responsible for comments made by third parties on their content online. However, and in line with the responses on this point in our consultation, in our view it is appropriate that comments on news articles are within the scope of the criminal offences we recommend.
4.34 We recommend that both the general harm-based communications offence and the knowingly false communications offence have an explicit press exemption. This exemption should not apply to comments made by readers on news articles online.
5.1 The nature of online communications has meant that existing legal frameworks addressing communications do not necessarily properly reflect the way people communicate. Three discrete areas of potential reform we considered in the consultation paper are group harassment, the glorification of violence and violent crime, and body modification content.
5.2 The phenomenon of uncoordinated group harassment, or “pile-on” harassment, has been the subject of much debate. The instant and geographically unbound nature of online communication is such that spontaneous harassment can exponentially concentrate on individuals. We considered the potential harm this, and coordinated group harassment online, can cause in the scoping report1 and the consultation paper.2
5.3 We asked two specific consultation questions and received many detailed responses. As we set out in Part 1 below, we have come to the conclusion that the most effective way of addressing this problem does not involve the creation of a further specific criminal offence. In our view, the combination of the application of existing criminal offences, which we consider in some detail, alongside robust non-criminal regulation and action by platforms will best address this problem. This approach will ensure that the most serious examples of abuse are recognised and punished without the risk of over-criminalisation or the further complication of criminal offences addressing expression.
5.4 Another area of online communication that formed part of our scoping and consultation was the potential glorification of violence and violent crime. We considered in detail the array of existing inchoate offences3 that cover inciting, encouraging, aiding, or abetting violent crime.4 The various harms that can arise from communications, including indirect causal harms, led us to seek consultees’ views on whether there should be an offence of glorification of violence or violent crime. We did
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, paras 8.160 to 8.222.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, Chapters 3,4 and 6.
Inchoate offences target conduct where a substantive offence may not be complete, but where what has occurred is worthy of a different type of criminal sanction. These include standalone offences of encouraging or assisting crime. For a more detailed discussion see paras 7.15 to 7.16.
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, Chapter 12; Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.152 to 174.
this as we were unconvinced that an offence targeting glorification of violent crime was a proportionate or appropriate response.
5.5 Part 2 below explores this in greater detail, in particular the range of responses we received that largely aligned with the concerns we raised in the consultation paper. Ultimately, we are of the view that an offence of glorification of violence or violent crime is not justified.
5.6 Lastly, we consider the ramifications of our recommendations in this Report for body modification content. The potential overlap between this content and any offences addressing glorification of violence or violent crime, or encouragement of self-harm was considered in the consultation paper and the subject of a specific consultation question.5
5.7 Group harassment online takes two forms: coordinated or uncoordinated.6 In the context of harassment by sending communications (the kind with which we are concerned in this part of the report), coordinated harassment occurs where multiple individuals act in concert to send communications that are harassing in nature to a victim. Uncoordinated “pile-on” harassment occurs where multiple individuals, acting separately, send communications that are harassing in nature to a victim. The instant and interconnected nature of online communications means that this phenomenon largely occurs in online environments.
5.8 An example that we described in the consultation paper was the abuse directed to Jess Phillips MP by hundreds of individuals with communications along the lines of “I would not rape you”.7 In our consultation, a number of consultees explained in meetings and in written responses that another type of group harassment that can occur is “dragging” or “trashing”, where groups of users target and abuse individuals.8 Another example is the ongoing and widespread online abuse of Premier League footballers, which involves both isolated harmful communications and spontaneous “pile-on” harassment.9 This type of harassment, as we noted in the scoping report and consultation paper, can have a serious impact, arguably greater than by an individual
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.200 to 6.207.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, Chapter 3; para 6.69ff; Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, para 3.78.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.71, citing M Oppenheim, Labour MP Jess Phillips receives ‘600 rape threats in one night’ Independent (31 May 2016), available at https://www.independent.co.uk/news/people/labour-mp-jess-phillips-receives-600-rape-threatsin-one-night-a7058041.html (last visited 13 July 2021).
See eg S Hughes, “Trolls on ‘dragging sites’ can ruin lives. It’s time they answered for their actions.” The Guardian (5 October 2020), available at https://www.theguardian.com/commentisfree/2020/oct/05/trolls-dragging-sites (last visited 13 July 2021).
See eg BBC Sport “Facebook ‘horrified’ by online abuse of Premier League footballers” (10 February 2021), available at https://www.bbc.co.uk/sport/football/56007601 (last visited 13 July 2021). We also received a Consultation Response from the Premier League which described in detail the abuse experienced by footballers and referees and their families.
given the “persistence and escalation” that would be difficult for an individual to replicate.352
5.9 The levels of abuse we discussed in the consultation paper have, sadly, continued to rise. Amnesty International’s reporting on the level of abuse that black women MPs faced for defending Black Lives Matter protests and the risk this poses to public and political life came against the backdrop of their other extensive work on online abuse:353
The persistent abuse women face on [Twitter] undermines their right to express themselves equally, freely and without fear. This abuse is highly intersectional and women from ethnic or religious minorities, marginalised castes, lesbian, bisexual or transgender women - as well as non-binary individuals - and women with disabilities are disproportionately impacted by abuse on the platform.
5.10 Online harassment continues to impact people across all sections of the community. As research by the Pew Research Centre illustrates, experiences differ. Their research indicates that women are more likely to encounter sexual harassment online, LGBT+ communities are particularly likely to encounter online abuse, and men are somewhat more likely than women to say they have experienced any form of harassment online.354
5.11 The COVID-19 pandemic has also exacerbated the incidence and impact of online abuse. The particular impact this has had on women and Black and minoritised communities was the subject of a report by Fix the Glitch and the End Violence Against Women Coalition, which explored the potential chilling effect of abuse and the increase in “piles-on” that has occurred.355
5.12 In some cases, existing provisions may capture group harassment. Coordinated harassment is covered by section 7(3A) of the Protection from Harassment Act 1997 (“PHA 1997”).
5.13 The group harassment offence in section 7(3A) PHA provides:
A person’s conduct on any occasion shall be taken, if aided, abetted, counselled or procured by another—
(a) to be conduct on that occasion of the other (as well as conduct of the person whose conduct it is) and;
(b) to be conduct in relation to which the other’s knowledge and purpose, and what he ought to have known, are the same as they were in relation to what was contemplated or reasonably foreseeable at the time of the aiding, abetting, counselling or procuring.
5.14 While the drafting is complex, it allows for flexible application. Ultimately, it means that the “course of conduct” criminalised by the PHA 1997 can be completed by two individual acts by two different defendants so long as the first defendant aided, abetted, counselled or procured the conduct of the second defendant. We also consider the potential applicability of this offence to those who aid, abet, counsel or procure conduct of more than one other defendant below, in the context of those who encourage or incite “pile-on” harassment but do not themselves participate in the harassment itself.
5.15 As we set out in the consultation paper and scoping report, this means that harassment perpetrated by a group can be prosecuted under the PHA 1997. For example, if the first defendant sent a harassing communication to a victim and then helped three other people draft harassing communications, which they individually sent to the victim, then the first defendant could be liable for harassment. By virtue of section 7(3A), the harassing communications sent by the other three people are deemed to be part of the first defendant’s course of conduct. This is because the first defendant aided (or abetted or counselled or procured) the conduct of the other three.356
5.16 The PHA 1997 does not provide protection for uncoordinated group (or “pile-on”) harassment. The lack of explicit coordination involved in the phenomenon means that it does not fall within the scope of section 7(3A) PHA 1997, which we noted was a gap in protection in the consultation paper.357
5.17 Individual communications that form part of an uncoordinated “pile-on”, while not covered by the PHA 1997, may amount to a communications offence. While it would depend on the content and context of each individual communication, it is possible that either the existing offences may apply, or, in line with our recommendations above, a harm-based communications offence.
5.18 The following extract from the consultation paper uses a series of examples to explore how the existing law applies to group harassment:358
Example 4: coordinated group harassment
A female MP, Ruby, recently made a statement calling for the close of the gender pay gap. Within 48 hours of making the statement, Ruby received thousands of emails, Tweets, and comments on her Instagram page. Some criticised her position on the gender pay gap. Others used misogynistic slurs, calling her a “stupid slut” or a “dumb bitch”. Some used the hashtag “#IwouldnotevenrapeRuby”. Ruby has been vocal on her own social media pages and in the press about the abuse she has received.
As it turns out, many of the Tweets were sent by people who were part of a 4chan thread created by Oscar. Oscar wrote: “We need to teach Ruby a lesson. The world should know what a dumb slut she is. Say it loud, say it clear! “#dumbbitch #IwouldnotevenrapeRuby... You know what to do.” Using an alias, he then Tweeted Ruby using the hashtags #dumbbitch and #IwouldnotevenrapeRuby.
6.78 Oscar would be likely to be guilty of the existing offence of harassment under the PHA 1997. This is because Oscar may be found to have counselled or procured the conduct of those who saw his 4chan thread and subsequently sent harassing Tweets to Ruby. Therefore, even though Oscar himself sent only one Tweet to Ruby, the Tweets of others who saw his 4chan thread must, under section 7(3A) PHA 1997, be taken to be part of his course of conduct.
Example 5: uncoordinated group harassment
Naomi, a young woman who follows Ruby on Twitter and Instagram, sees the harassing messages being sent to her and decides to join in. Underneath one of Ruby’s Instagram posts, she comments: “You stupid slut. She-devils like you need a good raping. but I wouldn’t want to be the one to do it #IwouldnotevenrapeRuby”.
6.79 Naomi’s comment would not be caught by section 7(3A) PHA 1997. This is because she was not part of and did not see the 4chan thread; she merely observed the pile-on and joined in. Unlike Oscar, she did not counsel or procure the conduct of other participants in the pile-on. However, Naomi’s comment would likely be caught by the current communications offences, and the proposed offence. Given the nature of Naomi’s comment, the prosecution may be able to prove that Naomi intended to cause harm amounting to at least serious emotional distress....
6.80 Regarding the conduct element, the comment was made in the context of a pile-on, which (as we have explained) can have a serious impact. We also note that the marginal harm caused by each individual communication does not necessarily diminish after a high volume of communications has already been received. Especially given that the comment was made in the context of a pile-on, the court should find that the comment was likely to cause harm amounting to at least serious emotional distress.
6.81 Finally, despite being made in the wake of Ruby’s statement on the gender pay gap, this particular comment was not a genuine contribution to political debate (and, it seems, was not intended as such): it was purely derogatory, and contained no remarks about Ruby’s political positions of any kind. It would not, therefore, be covered by the protection for political speech. It is highly unlikely that the court would find that Naomi had a reasonable excuse for posting the comment.
Example 6: political commentary
A political commentator, Mohammed, is not convinced by Ruby’s analysis of the gender pay gap. He is aware of the abuse she has received. He writes a Tweet directed to Ruby saying, “Good effort @RubyMP, but we need smarter thinking on this important issue” with a link to an article he wrote about the causes of gendered income inequality.
6.82 Mohammed’s Tweet would not be caught by the proposed harm-based offence. Mohammed did not have any intention to cause Ruby harm - instead, he wanted to point out the flaws in her analysis of the gender pay gap.
6.83 Mohammed may have been aware of a risk that Ruby would be caused harm (that is, emotional or psychological harm amounting to at least serious emotional distress): he sent a Tweet which was critical of Ruby and, arguably, patronising in tone in the context of a pile-on. Moreover, Ruby had been vocal about the abuse she has received and Mohammed knew about her experience.
6.84 Even so, Mohammed had a reasonable excuse. This is because all the facts suggest that his Tweet was a genuine contribution to a political debate and was intended as such. Even without the explicit protection for political speech provided under the proposed harm-based offence, a court would probably find that taking the risk of harming Ruby was not unreasonable, given the nature of the Tweet. With the explicit protection for political speech, it is clear that Mohammed’s Tweet should not be caught by the proposed offence: it was sent with reasonable excuse.
5.19 In our view there are two possibilities for addressing group harassment:
(1) Criminalising incitement or encouragement of group harassment; and
(2) Criminalising knowing participation in pile-on harassment.
5.20 In the consultation paper we set out our view that while there may be scope for the former, that there was not a good case for the latter.359
5.21 There are two forms of conduct which may not be adequately covered by either the existing law or our recommended harm-based offence. The first type of conduct involves a defendant who coordinates but does not participate in a pile-on. As we noted in the scoping report,361 and consultation paper,362 section 7(3A) seems rarely if ever to be prosecuted, even when defendants have themselves sent a harassing message, let alone when they have not. This is likely due to the complex formulation of section 7(3A), combined with the fact that it is a relatively low-level offence. We return to this when considering our consultees’ responses at paragraph 5.915.57 below.
5.22 It is also unclear that this conduct would be caught by the recommended harm-based offence. Communications coordinating a pile-on will tend not to be likely to cause serious distress to a likely audience: harm is generally caused to the target of the pile-on, not to the participants. In this situation, then, the participants in the pile-on could be prosecuted, but the instigator may escape criminal liability under the recommended harm-based offence.
5.23 However, a range of provisions exist that could address this. Section 8 of the Accessories and Abettors Act 1861 (“AAA 1861”) applies to any indictable offence - it would therefore include our recommended harm-based offence - and provides that anyone who aids, abets, counsels, or procures an offence is criminally liable as if they themselves had completed the criminal act.
5.24 The second, and related, type of conduct is unsuccessful encouragement of a pile-on. While unsuccessful encouragement would not be captured by the AAA 1861, it may constitute an inchoate offence. As we note in the scoping report, CPS prosecution guidelines suggest that “Those who encourage communications offences, for instance by way of a coordinated attack on a person, may be liable to prosecution under the provisions of sections 44 to 46 Serious Crime Act 2007.” 363The same would hold true for our recommended offence: someone who encourages the commission of the proposed offence may be liable under the Serious Crime Act 2007 (“SCA 2007”).
5.25 Under section 44 SCA 2007 it is an offence to do an act capable of encouraging or assisting an offence intending that it should be committed. Section 44 provides:
(1) A person commits an offence if-
(a) he does an act capable of encouraging or assisting the commission of an offence; and
(b) he intends to encourage or assist its commission.
(2) But he is not to be taken to have intended to encourage or assist the commission of an offence merely because such encouragement or assistance was a foreseeable consequence of his act.
5.26 Section 45 SCA 2007 concerns encouraging or assisting an offence believing it will be committed. Section 45 provides:
(1) A person commits an offence if-
(a) he does an act capable of encouraging or assisting the commission of an offence; and
(b) he believes-
(i) that the offence will be committed; and
(ii) that his act will encourage or assist its commission.
5.27 Section 46 SCA 2007 concerns encouraging or assisting multiple offences, believing that one or more will be committed. Section 46 provides:
(1) A person commits an offence if-
(a) he does an act capable of encouraging or assisting the commission of one or more of a number of offences; and
(b) he believes-
(i) that one or more of those offences will be committed (but has no belief as to which); and
(ii) that his act will encourage or assist the commission of one or more of them.
5.28 In theory, then, a person who unsuccessfully encourages a “pile-on” could be charged under one of sections 44 to 46 SCA 2007, with encouraging our recommended harmbased offence.
5.29 However, we are not aware of any prosecutions for encouraging - or aiding, abetting, counselling or procuring - the existing communications offences (regardless of whether the encouragement, or procurement, was successful in bringing about a communications offence). The complexity of such charges may be seen as disproportionate to the gravity of the encouraged offence. The proposed offence, like section 1 of the Malicious Communications Act 1988, would be an “either-way”364 offence.
5.30 The relative complexity of the above - in conjunction with what we have heard about the prevalence and harmful impact of pile-on harassment, combined with seemingly few prosecutions for such conduct365 - led us provisionally to conclude that there may be a case for a specific offence of inciting or encouraging group harassment. We asked a specific consultation question on the matter:366
Should there be a specific offence of inciting or encouraging group harassment?
5.31 A number of consultees supported the creation of a specific offence of inciting or encouraging group harassment. Many acknowledged our analysis of the various existing provisions that could apply to the behaviour, but stressed that a specific offence would make prosecutions more straightforward. Further, consultees emphasised the harm that this type of behaviour online causes.
5.32 The Magistrates Association agreed with our analysis that there are potential gaps that a specific offence could address: “...potential gaps in existing law that do not cover some of the examples given in the consultation paper.”367
5.33 The Alan Turing Institute agreed. They noted that “pile-on” abuse is often instigated by sophisticated perpetrators who “exploit and manipulate the algorithmic design of platforms” to amplify the abuse. They also noted the clear parallels with the Public Order Act 1986 “stirring up” offences:368
Co-ordinated group attacks can heighten the harm that is inflicted on victims by creating a sense of near-universal approbation. This often happens because, unfortunately, sophisticated actors exploit and manipulate the algorithmic design of platforms to amplify their activities. For instance, if 50 trolls worked together in concert then they could each send one original messages to a victim and also share and reply to the others. This would create close to 2,500 messages/replies. On most platforms, this would appear like an avalanche of abuse, yet it is the work of just 50 accounts, likely working for only a few minutes. Put simply, the ease of interaction and communication enabled by online platforms can be manipulated in this way to inflict substantial emotional distress on victims - far beyond what is inflicted in offline settings where such group behaviour is far harder and more costly to coordinate. Furthermore, in most cases there are far fewer individuals than accounts. In the example just given, 50 separate individuals might not be involved because multiple accounts could be controlled by a far smaller number of people (potentially even just one person), some of which might be bots or heavily automated.
As outlined in the Law Commission’s consultation report (pp. 155-161) there is a risk that the perpetrator/organiser of pile-ons may not be a message sender, and as such their actions would potentially not be covered by the existing law. This is problematic as we would argue that the harm is clearly being inflicted on the end recipient by the actions of the organiser, as much as by the message senders. In some cases, the causal link between the organiser and the message senders may be quite weak and such cases should not come under the purview of the law - we would strongly advocate that it is retained only for cases where an organiser knowingly and explicitly incited group harassment. This caveat aside, it is important that organisation of harassing behaviour is addressed by a new offence. We note that there is a clear precedent for this in how hate speech is tackled, where the 1986 Public Order Act criminalises actions to ‘stir up’ racial hatred. A similar view could be adopted here: the organisers are stirring up malicious communications.
5.34 Community Security Trust agreed and noted the severe impact of “pile-on” harassment on both the individual targets and the wider community to which they belong:369
The impact on the victim, and the wider community impact, is acute. Therefore, there is a need for a specific offence of this kind in order to mitigate the chances of this type of harassment not being prosecuted by older and more obscure legislation that could otherwise be used.
5.35 Refuge expressed strong support for a specific offence to address “pile on” and other forms of group harassment:370
Refuge strongly supports the creation of a specific offence of inciting or encouraging group harassment. ‘Pile on’ or other forms of group harassment can cause significant fear, harm and distress and often lead to survivors reducing their online presence, and therefore interferes with their rights to freedom of expression. The absence of such an offence is a significant gap in the law.
5.36 Fix the Glitch agreed, emphasising the disproportionate impact of coordinated collective online harassment on women and minoritised communities:371
Coordinated collective harassment is a growing problem online, and affects women and minoritised communities disproportionately. Research has shown that malign groups and individuals are using social media to launch coordinated harassment campaigns against public figures and social media users with no public profile alike. An investigation by BBC Newsnight in April 2019 showed that self-identified female politicians across Europe were targeted with threatening and misogynist content and coordinated harassment campaigns ahead of the European Parliamentary election. Research has consistently shown that women are more likely to be targeted by coordinated or uncoordinated harassment. In 2017, the European Institute for Gender Equality described cyber harassment against women and girls as a growing problem. Legislation in the UK has not kept up with this growing threat. Coordinated or collective harassment needs to be addressed as a specific offence.
5.37 Stonewall agreed with our analysis and noted the potential ease of prosecution that would be afforded by a specific offence:372
Stonewall supports the introduction of a specific offence of inciting or encouraging group harassment. Stonewall agrees with the Commission’s assertion that group harassment - whether coordinated or not - can have a significant impact on the victim(s) involved, both at the level of individual emotional and psychological harm, and a broader societal level, such as having a chilling effect on marginalised groups’ ability to participate in public life (6.73).
LGBT+ people often face harassment online and offline, includes instances that mirror the Commission’s description of ‘pile-on’ harassment. These events can have a significant impact on LGBT+ people’s ability to move through public life safely and free from harassment, causing many to retreat from online spaces.
5.38 Epilepsy Society agreed and provided evidence of “pile on” abuse the Society has faced on Twitter:373
Epilepsy Society agrees that there should be a specific offence of inciting or encouraging group harassment. The attacks Epilepsy Society, other epilepsy organisations, people with epilepsy, and the family and friends of people with epilepsy have experienced has often been as the direct result of inciting or encouraging group harassment.
During the attack on Epilepsy Society’s Twitter account in May 2020 users on Twitter stated ‘Let’s spam this to the epilepsy community’ and ‘thread of strobe light gifs to send to epileptix’. It was clear that this was a co-ordinated attack by a group of trolls deliberately setting out to harm people with epilepsy. Targeting people with disabilities should be considered reprehensible and Epilepsy Society would support this being reflected in the law and by further regulations on online platforms such as social media.
5.39 The Suzy Lamplugh Trust agreed, but noted that should culpable behaviour constitute stalking, the stalking offence should be used:374
Uncoordinated group harassment happens often in cases where the victim is in the public eye, such as in the case of MP Jess Philips or cases of Twitter trolling as seen within our Cyber Abuse as Work report. These cases are not currently adequately covered by law (notably where the perpetrator did not start the communication, such as adding a comment on a Tweet that causes harm) and it is vital that this changes. Our report put a spotlight on various forms of online abuse perpetrated in the workplace to devastating psychological effect. Among the victims of cyber abuse at work, a shocking 23% experienced trolling. These victims should be protected adequately under the law.
In cases of so-called ‘honour’ based violence it is common to see one member of the family encourage or incite others to perpetrate abusive and harmful behaviour towards another. Stalking can also be perpetrated in a ‘group’ setting, where stalkers encourage other individuals to stalk by proxy. This would equal coordinated group harassment as defined in the consultation.
Again, it should be noted that where the behaviours demonstrated constitute stalking, it is the stalking offence that should be criminalised first and foremost, with the communications offence only in addition and never replacing it.
5.40 Against the views of those in favour of a specific offence, a substantial number of consultees warned against the creation of new offences where existing provisions address the same behaviour. Many also stressed the logistical difficulties that the mechanics of “pile-on” harassment pose for the criminal law. Consultees pointed to the action that platforms can take to address the behaviour, and the non-criminal mechanisms that make up the detailed Government online safety strategy.
5.41 Professor Alisdair Gillespie disagreed with the need for a specific offence, noting the ability of the general communications offence to address the behaviour:375
I think this can be dealt with by the proposed offence, with the pile-on harassment being an aggravating factor.
5.42 The Bar Council of England and Wales argued that existing offences and ordinary sentencing processes, can sufficiently address the behaviour:376
We do not encourage the creation of such a specific offence: existing offences and powers of sentence upon conviction are sufficiently comprehensive and current to cover a form or forms of inciting and encouraging harassment where offending entails (whether or not targeting) a group or groups.
5.43 The Criminal Bar Association cited the “panoply of existing inchoate offences” alongside the general harm-based offence as capable of providing adequate protection:377
Intention to cause harm or knowledge that harm was likely to be caused and reference to the context would enable the new proposed offence to be used to combat group harassment. We do consider that guidance accompanying the offence should specifically address this form of conduct, as such behaviour is increasingly observed on social media platforms, but the new proposed offence and the panoply of existing inchoate offences should provide sufficient protection.
5.44 Sex Matters noted that the mechanics of online discussion and disagreement are not amenable to an offence of inciting or encouraging group harassment with sufficient nuance that also provides adequate protection for freedom of expression:378
We strongly disagree with these proposals. We do not think it is practical or proportionate to criminalise “being part of a pile-on”, which can simply be lots of people disagreeing with one person or an organisation. Nor do we think it practical or proportionate to criminalise “inciting or encouraging group harassment”, which can simply be quoting something someone has said and criticising it. We believe such a law would be an unacceptable infringement on free speech and impossible to police consistently, and that it would be used by organisations and powerful people receiving criticism to criminalise their detractors.
5.45 South West Grid for Learning (SWGfL) argued that any attempt to regulate “pile-on” harassment should be constrained, potentially via a “reasonable person” test to ensure that vexatious complaints do not shut down debate.379
5.46 The Adam Smith Institute outlined the various existing offences that could be used to combat “pile-on” harassment and opposed the creation of a new offence.380
5.47 The Criminal Law Solicitors’ Association characterised the difficulties in proof for this type of offending as “immense”:381
The Criminal Law Solicitors’ Association do not see that there is a specific need for inciting or encouraging group harassment. The difficulties which would be encountered in trying to prove such an action would be immense. The CLSA believe that this proposal is unnecessary particularly in light of the earlier proposals.
5.48 Fair Cop argued that practical implementation of any offence would be extremely difficult:382
It seems very unlikely there could ever be a successful prosecution of such an offence, given the inherent difficulty in establishing what had actually happened and who was inciting or harassing who.
5.49 ARTICLE 19 argued that the lack of a clearly defined offence of group harassment militates against any specific inchoate offence targeting the behaviour:383
In the absence of a clearly defined offence of group harassment (see our response to the next question), ARTICLE 19 is concerned that putting forward a related incitement offence is premature. We also query whether the issues raised by group harassment could not be better addressed through legislation or other policy measures on anti-discrimination or harassment in the workplace. In general, we would recommend any incitement offence to take due account of the treaty language of Article 20 (2) of the ICCPR and the Rabat Plan of Action to determine whether such incitement has indeed taken place.
5.50 The Free Speech Union observed that the fact that the potentially available offence (as pointed out in the consultation paper) is not currently prosecuted as insufficient evidence to justify the creation of a separate, specific offence:384
We see no compelling reason to create yet a further offence of inciting or encouraging group harassment here. As the Commission point out in Paras.6.87 -6.98, in most cases there is a possible charge already: the fact that prosecutors apparently do not use the available charges is in our view no reason to create a new criminal offence.
5.51 English PEN noted the damaging nature of group harassment online. They highlighted our analysis in the consultation paper that pointed to the array of existing provisions that could be used to combat group harassment. They also argued that there are practical measures that platforms can (and are) take to assist in combatting the phenomenon of “pile-on” harassment:385
We acknowledge the phenomenon of social media harassment and note that these activities involve, by definition, messages targeted at an individual who is likely to see them.
Although such harassment might be described as a form of expression (the participants are using social media, after all) we note that they also engage and interfere with the free speech rights of the target/victim. The ‘pile-ons’ render the social media platform unusable by the individual, and many are forced to avoid using the platform altogether.
However, we are wary about introducing a new offence to criminalise something which is very often impromptu and where the boundary between outrage and harassment is indistinct.
The consultation paper itemises a list of existing laws under which the organisers of group harassment might already be prosecuted. We suggest that these methods continue to be used to address the most serious offences.
Where a course of action is not already caught by an existing offence, we suggest that the law err on the side of freedom of expression, rather than introduce a measure that could encompass legitimate political organising. In general, we counsel against the introduction of new laws in order to capture ‘edge cases’ as seems to be the case here.
We note that technological innovation could offer sufficient protections to deal with the problem of online harassment. For example, Twitter introduced new features in 2020 to give users more control over who can reply to, or target a message at their account; and email software can be set up to filter abuse or unsolicited messages.
Terms of service and the forthcoming online harms regulations could also be deployed to deal with such issues, without involving the criminal law.
5.52 The Crown Prosecution Service noted the various existing provisions that could be used to prosecute group harassment, and stated that they were unaware of any matters where a prosecution has been prevented by the lack of an available offence:386
The existing law addresses incitement to group harassment under the S.8 of Accessories and Abettors Act 1861, s.7(3A) the Protection from Harassment Act 1997 and s.44 to 46 Serious Crime Act 2007...from a CPS viewpoint, we are not aware of any cases or situation where the police have referred the matter for a prosecution decision and where the unavailability of an offence of this nature has prevented a prosecution from taking place.
5.53 Carnegie UK Trust discussed the potential alternatives to criminal offences in dealing with “pile-on” behaviour. They also discussed “gossip forums” and the potential difficulty in determining when harassment occurs:387
Pile-ons are . harmful events that are fundamentally enabled by the prevalence of one to many and many to many services, as opposed to one to one POTS [“Plain Old Telephone Service”] at the time of the [1988] Act. The question remains whether they are something that should be dealt with in civil regulation biting on platform design (such as the Online Harms regime) or as an offence relating to an individual. An individual who organises or triggers one (as it can be one comment that becomes algorithmically promoted to a wider audience) might have some awareness of the harm that can be caused. In some instances, however, it may be that the platform - rather than being a neutral intermediary - constitute an essential motivating element in online speech and behaviours. An example of this can be seen in the context of gossip forums. The operators of some of these platforms have taken the decision to structure the forum threads by reference to particular named individuals. Depending, of course, on the content of the posts and their quantity, this could be harassment but what is central to this is the fact that the posts will, through the decision of the service operator, be focussed on a specific individual. Even if the person the subject of the thread does not see the posts (though this is unlikely), that person would still be a victim as it is likely that information concerning private aspects of the victim’s life have been shared (we assume here that sexual images or deepfakes are not involved). It may be that the offences under the Data Protection Act 2018 could be used, but this sort of environment could be seen as facilitating if not encouraging pile-ons, though the platform operator may not choose the individuals the subject of each thread.
5.54 Stakeholders were divided over the issue of incitement of group harassment.
5.55 Those who supported a specific incitement offence, (including the Alan Turing Institute, Fix the Glitch, Stonewall, the Judicial Security Committee and CST) tended to emphasise that group harassment online posed a particular form of harm. They stressed that this harm, though similar to that targeted by the communications offences more broadly, was particularly serious. Examples of victims being the targets of concentrated abuse powerfully speak to the damage that group harassment can cause. Further, as noted by the Alan Turing Institute (picking up a point we made in the consultation paper),388 those who co-ordinate this harassment do not always participate in the abuse itself. To that end, a specific offence would provide a separate and distinct mechanism to address the particular harm caused by group harassment, ensuring fair labelling and a straightforward route to liability.
5.56 As seen above, many of the consultees who disagreed included many of our legal stakeholders (including the Criminal Bar Association, the Law Society of England and Wales and the Criminal Law Solicitors’ Association). Each stressed the significance of the recommended harm-based offence, and noted that the existing avenues for criminalisation we discussed in the consultation paper could better address any incitement. They stressed the difficulties that could arise in proving offences, as well as the potential duplication and dilution of the significance of the newly proposed communications offences and existing harassment offences that may occur if a specific incitement of group harassment offence were enacted.
5.57 Reflecting on these responses has allowed us to reconsider one of our principal concerns in the consultation paper.389 Those consultees who raised various concerns with specific offences caused us to revisit the existing criminal offences that could potentially apply to this behaviour. In particular, that they may apply in a way that we did not fully contemplate in our consultation paper. We now consider that incitement or encouragement of group harassment where the person who incites or encourages the harassment does not personally participate can be addressed by application of the PHA 1997.
5.58 Where a person A encourages others to harass person B, and person B is then subject to a “pile-on”, but person A does not personally participate in the “pile-on”, the provisions of the PHA may well allow the conduct of the others “piling-on” to be attributed to person A.
(1) Section 7(3A) provides:390
(3A) A person’s conduct on any occasion shall be taken, if aided, abetted, counselled or procured by another—
(a) to be conduct on that occasion of the other (as well as conduct of the person whose conduct it is); and
(b) to be conduct in relation to which the other’s knowledge and purpose, and what he ought to have known, are the same as they were in relation to what was contemplated or reasonably foreseeable at the time of the aiding, abetting, counselling or procuring.
(2) If person A encourages or incites persons C, D and E to harass person B, and persons C, D and E then post communications “piling-on” to person B, then it would be open for prosecutors to prove that person A has aided, abetted, counselled or procured each of person C, D and E’s conduct harassing person B.
(3) For example, adapting Example 4 above, Oscar posts on 4chan, including a link to Ruby’s twitter account: “We need to teach Ruby a lesson. The world should know what a dumb slut she is. Say it loud, say it clear! “#dumbbitch #IwouldnotevenrapeRuby... You know what to do.” Other members of the 4chan thread Sam, Alex and David each separately reply to Oscar’s message saying they have done their part in “teaching Ruby a lesson”. Individually, Sam, Alex and David each Tweet to Ruby’s account with abusive messages employing the hashtags used by Oscar.
(4) Here, Oscar does not directly contact Ruby himself. However, his post providing the details of Ruby’s account, the hashtags and the call to “teach Ruby a lesson” are capable of amounting to aiding, abetting, counselling or procuring each of Sam, Alex and David’s subsequent communications.
(5) Section 7(3A)(a) PHA 1997 may then apply such that the communications “piling-on” to Ruby of each of Sam, Alex and David “shall be taken.to be conduct on that occasion of [Oscar]”.391
5.59 This would mean that where it can be proven that person A incites or encourages others to “pile-on” to person B, and a “pile-on” eventuates, even where person A does not personally participate in the harassment, they can be liable as a principal offender under the PHA 1997, or via the Accessories and Abettors Act 1861 for the recommended harm-based offence as we set out in the consultation paper. 392Further, should the “pile-on” amount to harassment where the victim is caused to fear violence will be used against them on at least two occasions, the “either way” offence under section 4 of the PHA 1997 is available and can lead to a maximum penalty of ten years’ imprisonment.393
5.60 We appreciate that the above is a relatively complex way of addressing incitement or encouragement of group harassment. However, we are conscious of the various and serious difficulties that consultees raised with any approach that seeks to criminalise the behaviour by way of a specific offence. Proof of causation is a particular problem in this context, whether using existing offences or a new specific offence. In our view, and as consultees set out, that issue will continue to be problematic and the creation of a specific offence will not overcome it. Causation will be a matter requiring detailed consideration in each case on its facts. We also are particularly conscious that the Crown Prosecution Service was unable to point to any matters where existing provisions could not apply to address culpable behaviour.
5.61 Further, we acknowledge that this approach relies on there being conduct that has been “aided, abetted, counselled or procured” by person A to establish liability. The situation we discussed in the consultation paper where there is unsuccessful encouragement of group harassment would only be covered by way of inchoate liability. 394As we noted in the consultation paper this may be seen as a disproportionate approach to an “either-way” offence by police and prosecutors. However, as we noted above, the relevant prosecutorial guidance explicitly contemplates inchoate offences being used where individuals encourage communications offences by way of a coordinated attack on a person.395
5.62 Some of the difficulties that we outlined with this approach remain: proving intention or belief as to a “pile-on” may be challenging. However, we note that the only area the existing law may not adequately cover relates to instances of encouragement or incitement where no “pile-on” has eventuated. That is, no harm to the intended target has been caused. The most serious instances of group harassment online are covered by the existing offences. The case for adding to an already complex area of the criminal law, or taking the risks that attend further criminalisation that consultees pointed out in their responses,396 is therefore weak.
5.63 Other consultees, including the Carnegie UK Trust, noted the potential ability for Online Harms non-criminal regulation to address the mechanics of “pile-ons”. They queried whether, given the nature of online communication, a specific inchoate criminal offence was the appropriate way to address the behaviour. The combination of the requirement for companies to ensure illegal content is removed expeditiously397 with the regulation of “legal but harmful” content online 398militate against enacting a further specific criminal offence.
5.64 A further issue arose as a theme in consultation responses: where a “pile-on” is instigated by a person who either does not intend for it to eventuate, or is in fact attempting to encourage others to send messages of support to a third party. The PHA 1997 offence requires that the perpetrator “knows or ought to know involves harassment”, 399where harassment includes “alarming [the victim] or causing [the victim] distress.” 400As a result, the provisions ensure that only intentional encouragement or incitement of group harassment would be captured.
5.65 Many consultees who repeatedly stressed the importance of freedom of expression, for example the Free Speech Union, the Open Rights Group and ARTICLE 19, were also concerned about the creation of a specific incitement offence on freedom of expression grounds. While their concerns about the nature of such offences have already been made apparent elsewhere (in that they broadly oppose criminal regulation of speech), they also noted that an incitement offence would exacerbate their concerns given its further distance from harm, and the applicability of the existing provisions. We appreciate that the creation of a specific offence, in circumstances where the conduct sought to be addressed is already substantially or wholly covered by existing offences, could be seen as an unreasonable interference with freedom of expression.
5.66 Our analysis of the existing offences that could potentially be used to address incitement of group harassment was echoed by both the Crown Prosecution Service and English PEN in their responses. Other stakeholders also noted that given impugned behaviour could be addressed adequately with existing offences, the creation of a new offence could create further uncertainty, and that the fact that an offence has taken place as part of a “pile-on” can be taken into account at sentencing.
5.67 After considering the detailed consultation responses we have received, we are not persuaded that a specific offence of inciting or encouraging group harassment is appropriate, or needed.
5.68 We are sympathetic to the views of our stakeholders and consultees who would prefer the matter be dealt with by way of a specific offence. We understand the importance of tackling group harassment online, and believe that the articulation of the varying existing offences that we undertook in the scoping report and consultation paper, and outlined again above, is important in ensuring that law enforcement is aware of the scope and potential application of the existing criminal offences on the matter.
5.69 However, ultimately, we think that the combination of direct action by platforms and the important work of the online harms regulator with the rigorous application of existing offences is the most appropriate way of addressing incitement or encouragement of group harassment online.
5.70 The second option for reform was a specific offence targeting knowing participation in “pile-on” harassment. This would attempt to address situations where a single person sends a single abusive communication that does not by itself meet the criminal threshold (such that the recommended harm-based offence does not apply):
(1) in the knowledge that similar abuse is being targeted at the victim; and
(2) with an awareness of the risk of greater harm to the victim occasioned by their conduct in the circumstances.
The various challenges in creating an offence to meet these circumstances were the subject of our analysis in the scoping report401 and consultation paper.402 These include evidential challenges relating to each of the above conditions, and the lack of applicability of the PHA 1997 to uncoordinated conduct.
5.71 These concerns led us provisionally to conclude that there was insufficient justification for a specific offence of knowing participation in “pile-on” harassment. We asked a specific consultation question on the topic:403
Should there be a specific offence criminalising knowing participation in uncoordinated group (“pile-on”) harassment?
5.72 Fix the Glitch argued that the seriousness of “pile-on” online abuse is such that it should be addressed with a specific offence:404
Group harassment is a particular form of online violence whose seriousness and coordinated nature needs to be acknowledged in legislation. Sometimes, however, collective harassment can grow organically, fuelled by social media’s and online platforms’ amplifying effect. Recent high-profile cases of collective harassment online have shed light on the amplifying effect of social media in fuelling harassment. In March 2020, a video of a hospitalised COVID-19 patient originally shared via the messaging application WhatsApp was shared to social media and fuelled a large scale collective abuse with collective harassment spreading organically rather than in a coordinated way.
5.73 Dr Kim Barker & Dr Olga Jurasz argued in favour of a specific offence to address “pile-on” harassment as part of a re-conceptualisation of laws to combat online abuse:405
The creation of a specific offence of ‘pile-on’ harassment would be welcome. Such legislative development would allow for further conceptualisations of online harms, which takes into consideration the reality of online abuse, its scale (including cross-jurisdictional spread), the speed with which it is committed, and - most importantly -the impact it has on the victim(s). In our research, we have criticised the current efforts to reform the law on online harms as representative of an incredibly fragmented and selective approach whilst advocating for a conceptually broader understanding of online harms. We would strongly support the amendment of the law in this direction.
5.74 Professor Tsachi Keren-Paz argued that while the general communications offence will, by and large, address much of this behaviour, the distinct harm inflicted by “pile-on” abuse is such that a separate offence could be justified:406
The stakes are not enormously high here, since I agree that many of these communications would be caught by the general suggested offence as likely to cause serious emotional distress. However, to the extent that each alone is not likely to cause harm (especially if the bar to the offence remains serious harm) but the defendant is aware (and I would add or ought to be aware) that it is likely that many others will send similar messages and that the effect on the victim would hence be more serious, there is a strong justification to impose criminal responsibility: Once it is accepted that pile-on abuse is additive and hence is more serious, knowing participation in it amounts to an anti social behaviour which ought to be caught by the criminal responsibility net.
5.75 The Criminal Bar Association warned against criminalising participation in uncoordinated “pile-on” abuse. They noted the ability of the proposed communications offence to address much of the behaviour:407
...if such activity is charged under the new proposed offence the court should take into account when considering the facts of such a case that there may be in fact greater [culpability] attached to those who send messages later in time in a pile-on as they had greater opportunity to be aware of the harm likely to be caused in the circumstances of them adding to the number of messages/posts already sent/posted; a pile-on by definition becomes more harmful the longer it goes on and the more messages (/people) are involved.
The courts could find that the relationship between the defendants’ messages was ‘similarity of impact’ or ‘similarity of [general] intent’ - e.g. defendant 1’s message may be homophobic, defendant 2’s message may be fat-phobic, in the context of a pile-on both messages are capable of being harmful (similarity of impact) and are being sent with the general intent to.cause distress/cause harm etc (similarity of intent).
5.76 The Association of Police and Crime Commissioners did not support a specific offence, noting that “Communications sent during a pile-on which reach the threshold of causing “serious emotional distress”, will be covered by harm-based offence as outlined above.”408
5.77 The Magistrates Association noted the issues with culpability that may arise in any offence targeting uncoordinated “pile-on” abuse. They also recognised that the general communications offence may be able to address this type of behaviour:409
We note the explanation set out in the consultation about complications in evidencing a separate offence for “pile-on” harassment. We agree that even though the harm caused by such actions can be great, it may be difficult to link culpability of one individual sending one message to that greater harm, unless they were involved in inciting, encouraging or otherwise coordinating the group action.
However, we do note that if someone was aware that there was a “pile-on” occurring, and chose to add to it by sending a message that was otherwise covered by the new proposed harm-based criminal offence, then possibly this could be regarded as an aggravating factor. If there were evidence that someone was responding to a message from the victim (for example) indicating they were receiving hundreds of abusive messages, possibly linked to a specific hashtag, then this might indicate greater culpability for more serious harm caused.
5.78 The Alan Turing Institute noted the potential for over-criminalisation posed by any offence that aimed to address uncoordinated “pile-on” abuse:410
We recognise the positive and proactive motivation behind this proposal, and the need to adopt a victim-centric approach which accounts for the total impact of malicious online behaviours on those who are targeted. However, our concern is that if this offence is addressed at “uncoordinated groups” then it may end up by criminalising the actions of individuals which, by themselves, would not be criminal. Instead, actions of one individual could become criminal if, in concert with many others, they lead to an accumulation of harmful behaviours. Whilst the impact on the targeted person may be highly harmful - and some form of appropriate intervention should be taken, potentially by the host platforms - it seems unfair to assign blame to each of the perpetrators for the actions of other individuals.
Participating in an uncoordinated ‘pile-on’ may be undertaken without awareness of the pile-on taking place; an individual may unwittingly have contributed to a pile-on which they did not know about, or want to be associated with. It seems particularly unfair for them to receive a heightened sentence in such a situation. Relatedly, we also caution that the definition of a ‘pile-on’ seems unclear and may be liable to misinterpretation and, most worryingly, an overly broad definition of what constitutes a pile-on. The law could easily be used in an overly draconian way to criminalise nearly any interpersonal aggression if it formed a (potentially quite loosely defined) pile-on.
Finally, we note that whilst we are concerned about uncoordinated group harassment being treated specifically under law, we support efforts to address organised networks of abuse and harassment.
5.79 The Bar Council of England and Wales noted that this may be “overly inchoate” behaviour to address with criminal sanction:411
We support introducing the new proposed offence considered in Chapter Five offence [to replace section 127(1)]. Knowingly participating in uncoordinated group (“pile-in”) harassment is overly inchoate in comparison to the preferred new offence proposed in Chapter Five.
5.80 The Crown Prosecution Service discussed the potential difficulty of proving a defendant’s awareness or knowledge of others’ actions and motivations:412
We note the Law Commission’s observations that such an offence may be difficult to enforce for the practical reasons outlined. There may be evidential challenges in demonstrating that individuals were aware of communications sent or posted by others, in distinguishing the culpability of the first instigating communication from subsequent communications, or to link them. There may also be numerous communications from different individuals making it resource intensive to enforce.
5.81 Demos agreed that “pile-on” abuse is a phenomenon requiring attention given the impact it can have on both targets and onlookers. They went on to provide a detailed response describing the nature of “pile-on” abuse online, and questioning the ability of criminal laws accurately and reliably to address the behaviour:413
The phenomena of pile-on or uncoordinated group harassment certainly deserves attention, due to the significant negative impact it can have on its targets as well as on onlookers. We welcome the examination by the Law Commission on action which can be taken to reduce group harassment. As highlighted by Glitch, a UK charity which raises awareness of online abuse, group harassment is a significant concern: it combines attacks on individuals, which can cause significant emotional distress, with communication to onlookers as another audience who may also suffer distress. For instance, women and members of other marginalised groups who observe others being attacked by groups online may themselves feel less able to engage in online spaces for fear of suffering the same targeting.
However, understanding and debate over pile-on harassment is currently impoverished. Some of the complicating factors for straightforwardly addressing pile-on harassment are as follows.
Firstly, pile-ons can include both coordinated and uncoordinated efforts, and many cases can lie somewhere in the middle with a small group privately and deliberately planning coordinated harassment elsewhere, often with a view to presenting the initial activity as authentically uncoordinated and with a view to encouraging others uninvolved in the initial plans to organically participate. This porousness means distinguishing coordinated from uncoordinated group harassment online can be extremely difficult: it is not always clear to someone engaging in a pile-on that they are doing so.
This has a bearing too on the standards by which we judge someone’s ‘knowing’ participation. The degree to which a participant can be said to know that the harassment is coordinated or not is similarly difficult to establish for the same reasons. Indeed, the degree to which they can be said to know that their particular online communication is part of a broader group activity is difficult to ascertain too. Not least because, given the rapidity with which many users engage with content online and thus give little consideration to context as a result, it may not occur to them that they are contributing to an activity with a mass of participants. This argument, of course, could be used disingenuously as a defence, but therein again lies the difficulty of establishing facts about pile-on harassment.
Two further complicating factors stem from the fact that it can occur over extended periods of time. On the one hand, this further complicates all of the above: for example, the sheer amount of data that can accrue around a hashtag used specifically in a pile-on can quickly exacerbate the difficulties of establishing its origins, who was involved at the start, the awareness of its un/coordinated status to participants and so on. On the other hand, the longer a pile-on continues, the more attention it is likely to attract, and so the more it is likely to be subverted by those who oppose it. In June 2020 many members of the online community of ‘K-pop stans’ (fans of a genre of South Korean pop music) co-opted the hashtag #whitelivesmatter (a hashtag developed deliberately in an online far-right community to undermine the anti-racist hashtag #blacklivesmatter and as a vehicle for far-right propagandising), in response to its increased use by those opposing the protests against the killing of George Floyd in the United States. The K-pop stan users deliberately rejected this use, instead accompanying it with expressions of support for those protesting. A similar example can be found in the hijacking by the LGBT community of the #ProudBoys hashtag in rejecting white nationalism. This example adds to the mix of difficulties discussed above, the factor of delineating when mass group activity online constitutes pile-on harassment.
5.82 Refuge noted that while supported in principle, any offence targeting knowing participation in uncoordinated group harassment may be unworkable in practice.414
5.83 Kingsley Napley discussed the potential practical difficulties that would meet any attempt to address knowing participation in uncoordinated group abuse:415
It is not clear how an individual knowingly participates in an uncoordinated group. This might lead to the investigation of someone who is unaware of third party communications and/or is not quick enough to realise that they could be considered a ‘pile on’.
5.84 ARTICLE 19 raised concerns about the impact that any offence targeting uncoordinated online harassment may have:416
ARTICLE 19 recognises that ‘pile-on’ harassment is a significant issue for public debate online and that it tends to target particularly vulnerable or marginalised groups or individuals. However, we are concerned that such an offence would criminalise individuals getting caught for making a one-off offensive comment that would otherwise fall below the threshold for prosecution of a harassment charge. In particular, it is unclear to us how an individual would be supposed to acquire knowledge of participation in ‘uncoordinated group harassment’ and the point at which the threshold for such knowledge would be reached, i.e. how many comments would be necessary for ‘uncoordinated harassment’ to take place? It is also unclear what kinds of tweets and how many would be sufficient to amount to ‘harassment’. What should be the relationship between the tweet or comment that sparked a ‘backlash’ and the tweets or comments in response and how much should the various tweets in response correlate with one another to amount to ‘uncoordinated harassment’? More generally, we are concerned that this offence could have a chilling effect on matters of public debate where people may strongly disagree for fear that one comment, however minor, may be taken to add on to ongoing harassment of someone online.
ARTICLE 19 reiterates that the criminal law should only be used as a matter of last resort over other policy responses. In the case of “pile-on” harassment, other measures such as improved content moderation practices, restrictions on certain accounts or features, would likely be more proportionate and appropriate.
5.85 Sex Matters set out their strong opposition to any proposal to criminalise knowing participation in “pile-on” abuse:417
We do not think it is practical or proportionate to criminalise “being part of a pile-on”, which can simply be lots of people disagreeing with one person or an organisation. Nor do we think it practical or proportionate to criminalise “inciting or encouraging group harassment”, which can simply be quoting something someone has said and criticising it. We believe such a law would be an unacceptable infringement on free speech and impossible to police consistently, and that it would be used by organisations and powerful people receiving criticism to criminalise their detractors.
5.86 While there was strong support for addressing communications that are intentionally harmful in the context of “pile-on” abuse, a majority of consultees expressed concern in doing so with a specific offence.
5.87 Some consultees were in favour of a specific offence. In particular, stakeholders who work extensively with victims of online abuse noted the acute harm that can be inflicted by a “pile-on” online. We have reflected at length on the various ways to address the behaviour. This has included whether a specific offence could be crafted that could be appropriately constrained and directed, as well as being easy to understand and straightforward to apply in practice. Unfortunately the relative strengths of a specific offence were unable to outweigh the potential for overcriminalisation. Similarly, the problems relating to proof (including causation) that arise in relation to existing offences would remain for any specific offence.
5.88 Conversely, a majority of consultees expressed concerns with the practical operation of any offence, and the potential over-criminalisation that any attempt to address the behaviour may entail given the nature of online communications. Many stressed that the general communications offence would address culpable behaviour, and that (in the ordinary course) the circumstances of the offending, including that it may or may not have occurred in a “pile-on” could be taken into account in sentencing. As for the general and false communications offences, consultees also expressed their concerns with the impact of any offence on Article 10 rights.
5.89 We acknowledge the harm that uncoordinated “pile on” harassment can cause. As set out above in relation to incitement of group harassment, we understand the concerns of stakeholders who would prefer this behaviour be dealt with by way of a specific criminal offence.
5.90 However, the detailed responses provided by consultees have only confirmed our concerns as to the workability of and justification for a specific offence of knowing participation in “pile-on” harassment. These responses supported our provisional conclusion that there is insufficient justification for a specific offence. 418We therefore confirm that conclusion that the combination of direct action by platforms and the important work of the online harms regulator with the application of our recommended harm-based offence (where the context of any communication being part of a “pile-on” may be taken into account, including as a possible aggravating factor at sentencing) is the correct and proportionate way to address knowing participation in “pile-on” harassment.
5.91 We considered in some detail the array of existing inchoate offences covering inciting, encouraging, aiding or abetting violent crime in the scoping report. 419Despite these, in the consultation paper we noted the potential to create specific offences that address “glorifying” violence or violent crime, and sought consultees’ views on the matter.420
5.92 A range of online content has been discussed in the context of glorification of violence. Perhaps most prominent is Drill music, “a genre of rap music that has proliferated in London since the mid-late 2010s” 421and associated videos on sites like YouTube.422 As we discussed in the consultation paper, there are some examples of existing mechanisms 423being used by law enforcement to address similar content, yet there are real concerns that the association of drill music with the glorification of physical violence is misguided.424
5.93 A report published in February 2021 by the law reform and human rights organisation JUSTICE examined racial injustice in the youth justice system. The report considers Drill music in some detail. It argues that the use of Drill music in the criminal justice system reveals a “myopic view” of cultural expression and assumes those associated with the creation of Drill music represent an inherent danger. 425They also noted the disproportionate impact that the focus on drill music as evidence, particularly when
used as evidence of bad character, has on Black men and boys.426 Of course, where directly relevant to a specific crime, a piece of Drill music may rightly be admitted as evidence. 427The central concern of those who urge caution in the treatment of drill music is the “...corrosive effect of portraying a genre of music so closely connected to Black communities as innately illegal, dangerous and problematic.”428
5.94 We discussed other instances of glorification of violence or violent crime in the consultation paper, including its application in the context of the death of George Floyd - a 46-year-old black man who died while being arrested in the US city of Minneapolis, Minnesota when a police officer knelt on his neck429 - and subsequent protests. A Tweet by then-President Donald Trump, in which he suggested that people looting in the wake of George Floyd’s death would be shot, was found by Twitter to violate the platform’s rules against glorifying violence.430 Further, after the storming of the United States Capitol in Washington DC on 6 January 2021, Twitter permanently banned Trump “.due to the risk of further incitement of violence”.431 It is noteworthy that Twitter’s definition of “glorification” has an extremely broad scope, though seems narrower than a definition that incorporates simple praise of violence, suggesting variance and ambiguity in the definition of the term.432
5.95 Where content encouraging violence (including gang violence) constitutes an act capable of encouraging or assisting the commission of an offence, and the person who posted the content intends to encourage its commission, it may be caught by the offences under sections 44 to 46 of the Serious Crime Act 2007 (“SCA 2007”).433
However, without an intention or belief that a criminal offence will be committed, the offences under the SCA 2007 will not apply.
5.96 As we discussed in the consultation paper and mentioned above, some content which police consider “glorifies” violence may be the subject of Criminal Behaviour Orders, the breach of which can lead to a prison sentence.434 Further, the offences in the Terrorism Act 2000 employ a broad definition of “terrorism” which includes “serious violence against a person” and “serious property damage”, endangerment of life, creating a serious public health or safety risk, and serious interference with electronic systems.435 Lastly, the recommended harm-based communications offence may well capture certain content that glorifies violence or violent crime: in particular, where a communication is likely to cause psychological harm to a likely audience that amounts to at least serious distress.
5.97 We also noted the possibility that alternative reforms may be able to address similar conduct without the wide-ranging impacts on freedom of expression. Namely, our 2015 recommendations to expand the offence in section 16 of the Offences Against the Person Act 1861 (“OAPA 1861”) from threats to kill to include threats to rape and threats to commit violence 436and our recommendations in Chapter 3 regarding threatening communications.
5.98 The example and discussion from the consultation paper sets this out:437
Example 9: glorification of violence causing serious emotional distress
Freddie sends video clips of a woman being violently attacked to an acquaintance of his, Sara. Both Sara and the woman in the video clips are black. Along with the clips, Freddie sends an accompanying message “Could have been you, Sara! They really got her good.” The message includes a string of “thumbs up” emojis.
6.164 Recall that, under existing legislation, “glorification” means any form of praise or celebration. Given that Freddie’s accompanying message said, “They really got her good” and included “thumbs up” emojis, his conduct may amount to “glorification” of violence.
6.165 This “glorification” of violence would likely be caught by our proposed offence. A court would probably find that Freddie’s communications were likely to cause Sara at least serious emotional distress. Further, the prosecution could probably prove that Freddie intended to cause harm or was aware of a risk of causing harm. There is nothing to suggest that he had a reasonable excuse.
5.99 In the consultation paper we set out our concerns with the potential vagueness of any offence that specifically targeted the glorification of violence or violent crime.438 In particular, we noted that while the offences that target the glorification of terrorism under the Terrorism Act 2006 may be justified as necessary in a democratic society given the especially acute threat that terrorism poses to society as a whole, those offences have been the subject of sustained criticism by human rights and civil liberties organisations.439
5.100 We provisionally concluded that there was insufficient justification for a specific offence of glorification of violence or violent crime. We sought consultees’ views on the matter with a specific consultation question:440
Should there be a specific offence of glorification of violence or violent crime? Can consultees provide evidence to support the creation of such offence?
5.101 Some consultees noted the prevalence of calls to violence online, and the potential for a specific offence to assist in combatting this. A number of consultees responded with specific concerns about the glorification of antisemitic violence,441 and graphic threats of violence toward women.442
5.102 Fix the Glitch agreed that a specific offence could be of use, noting their observations of an increase in calls to violence online:443
We welcome this suggestion. Over the last few years, we have seen an increase in calls to violence online, which can overlap with threatening communications. Research by the HateLab at Cardiff University in 2019 recorded a significant increase in overt hate speech online in the UK. Politically or ideologically motivated groups are increasingly using social media platforms to issue calls for violence against specific communities or groups of individuals, or promote gang violence. Researchers have, in particular, shown that recent months and the COVID-19 pandemic had led to an increase in calls for violence online and fuelled online extremism. The UK Commission for Countering Extremism showed that extremists across the political spectrum have exploited the pandemic and increased time spent online to sow division and encourage extremist agendas with a potential to fuel violence.
5.103 The National Police Chiefs’ Council were in favour of a specific offence being enacted. The views offered in their response included:444
There was general support for this although not a significant amount of evidence of individual crimes. Given the diverse views, I have selected below those that raise distinct issues:
1. We are not aware of this being a significant problem. We have had an issue with videos from [Organised Crime Group] members similar to those described but feel that proving the required fault elements would be extremely difficult. We agree that it would be better to have a general offence under the OAPA regarding threats to commit violence and rape.
2. The Serious Crime Act 2015 has similar legislation for gang injunctions, which we have used successfully in Essex.
The court has to be satisfied on the balance of probabilities that the respondent has engaged in or has encouraged or assisted—
(a) gang-related violence, or
(b) gang-related drug-dealing activity.
So we have legislation that allows us to target person(s) encouraging gang violence, i.e. drill music. But the lines are very close in terms of the cross over to films and other forms of media that also portray and glorify violence. It would need some care consideration to deconflict this. But I would be supportive of some stand-alone legislation wider than just Gang Violence.
5.104 Kingsley Napley noted that the existing offences, and provisionally proposed communications offences, likely provide sufficient protection:445
Glorification is probably sufficiently covered already but this area should be kept under review. Consideration would need to be given to the following aspects:
- Whether this would be potentially covered by other offences for example Terrorism Act offences.
- Whether it is possible to consistently and properly identify inappropriate behaviour ie. how comments made in jest would be excluded from criminalisation for example tweeting a meme which jokes ‘I am going to find you and I am going to kill you’
5.105 Dr Jen Neller argued an alternative way to address the behaviour would be by amending the existing encouragement offences:446
This could be dealt with more effectively by amending the Serious Crime Act 2007 to clarify the encouragement offences and extend them to encouragement of the public or any sector of the public to commit a crime, rather than only encompassing encouragement of a particular individual. Encouragement to commit a crime would be a higher threshold than glorification, and would avoid the risk of disproportionately criminalising black cultural expressions or of otherwise interfering with the freedom of expression of those who have grown up in or live in violent contexts (see Kubrin, Charis E. and Erik Nielson. “Rap on Trial.” Race and Justice 4, no. 3 (2014): 185-211). The further criminalisation of such marginalised populations would be at risk of causing greater harm than it remedied. Additionally, the criminalisation of glorifying violence would seem to inappropriately place a greater restriction on the expression of individuals than on the media (e.g. violent films and video games).
5.106 The Magistrates Association agreed with the views expressed in the consultation paper, particularly the concerns about Article 10:447
The arguments put forward in the consultation paper against any such new offence seem strong. We agree that any definition of glorification would have to be very tightly drafted, so that it did not infringe Article 10 rights. For example, we would be concerned about any offence that covered all drill music by default, for the reasons put forward by the Law Commission. It also seems clear that glorification that reaches the level of threats would be covered in existing legislation.
5.107 ARTICLE 19 argued that a “glorification” offence would be “very dangerous” for freedom of expression. They noted that no matter how an offence is drafted, the inherent ambiguity of “glorification” means that disproportionate “chilling” of expression is, in their view, inevitable:448
In our view, such an offence would be very dangerous for freedom of expression. Under international standards on freedom of expression, ‘glorification’ is generally considered too broad a term in relation to incitement to commit acts of terrorism or incitement to discrimination, hostility or violence (Article 20 ICCPR). Moreover, the glorification of ‘violent crime’ could potentially cover a very wide range of conduct that most people would not know are criminalised. It is unclear how the proposed offence would apply in several scenarios. For instance, would supporting violent police action fall within scope? What about the actions of the military which are highly likely to involve violence? Slavery is clearly an affront to human dignity. Some may argue it is a violent crime. Would those who seek to justify it fall within scope? We also note that a specific offence of glorification of violence could have an impact on minority groups, for instance those individuals whose sexual preferences include bondage and other sado-masochistic practices. Finally, we draw attention to a French case where a mother and an uncle were fined several thousands of euros for a similar offence in circumstances where the mother’s son, named Jihad and born on 11 September, had been wearing a T-shirt saying ‘I am a bomb’ at the front and ‘born on 09/11’ at the back given by his uncle for his 3rd birthday. In our view, the prosecution and conviction in this case were clearly disproportionate and demonstrate how such offences have a chilling effect on freedom of speech.
5.108 The Criminal Bar Association argued against any offence of glorification of violent crime. They discussed the potential disproportionate impact any offence may have on communities or sub-cultures:449
No; we believe that an attempt at criminalising behaviour under this banner will disproportionately affect some communities or sub-cultures (as per the e.g. of [Black] youth who listen to or themselves create “drill” style rap). We do feel that the association of drill music with the glorification of physical violence is misguided and racially biased and we not[e] that there are other genres of music which also have lyrics speaking of violence which are not viewed in the same way (e.g. hard rock).
In the field of art, music and artistic expression, the question of what in fact amounts to “glorification” of violence is often determined on subjective factors (e.g. the perception that drill music is fuelling violence in inner cities or is more lyrically graphic then other forms of music) this subjectivity we feel can lead to inequality in the way the law is applied to some artforms / music leading to further disproportionate criminalisation of some communities / cultures.
5.109 English PEN discussed their concerns about Criminal Behaviour Orders being used to prevent the performance of certain music. They agreed with our analysis that “street illiteracy” can lead to disproportionate responses to certain types of music and expression:450
We are extremely concerned by the recent prosecutions of rappers for producing ‘Drill’ music, and the Criminal Behaviour Orders preventing others from performing such music. We consider these interventions a disproportionate interference with freedom of expression rights.
We welcome the Commission’s acknowledgment in the consultation paper that such music and associated videos are susceptible to misunderstanding and ‘street illiteracy’ and that such content “represents, rather than endorses violent crime.” Such representation is a window into a segment of British society that is often marginalised, and is therefore in the public interest. It should be contextualised, not be criminalised.
We support the Commission’s implied suggestion a[t] paragraph 6.159 that only communications/videos that meet the threshold of incitement to commission of criminal conduct should be criminalised. Where no intent to incite a specific offence can be shown, and where the communication is not captured by the proposed harmbased offence (paragraph 6.163), such messages should not be prosecuted.
We support the points made in defence of freedom of expression at paragraphs 6.170-6.174 of the consultation paper, questioning the efficacy and proportionality of a ‘glorification’ offence.
5.110 The Crown Prosecution Service agreed that an offence of glorification or incitement of violence or violent crime may lack clarity, distinguishing the offences under the Terrorism Act 2006. They discussed the offences under the Terrorism Act 2006 and the “very specific” nature of terrorism and why the general concepts of “violence and violent crime” are not amenable to similar regulation:451
The Law Commission considers that a specific new offence of glorification or incitement of violence or violent crime may be too broad and lacking in clarity. We consider that there is merit to that analysis. This proposal can be distinguished from incitement offences under the Terrorism Act 2006.
The Terrorism Act 2006 includes references to glorification, defined as ‘direct or indirect encouragement or other inducement to some or all of the members of the public’. The Counter-Terrorism Division of the CPS has considerable experience of dealing with criminal glorification in the context of terrorism, in particular by way of the offences under Part 1 of the Terrorism Act 2006. The offences, which criminalise the encouragement of terrorist acts, have been an extremely useful tool in the fight against terrorism and have allowed investigators and prosecutors to successfully target those involved in the radicalisation of others.
Indirect encouragement under the 2006 Terrorism Act can form the basis for a criminal prosecution provided that other criteria are met. The additional criteria have the important effect of narrowing the scope of the offences and hence addressing some of the concerns, highlighted in the Commission report, relating to freedom of speech. One of the key criteria is that members of the public should reasonably be expected [to] understand that the terrorist acts are being glorified as ‘conduct that should be emulated by them in existing circumstances’ (Part 1 s.3 Terrorism Act 2006).
However, Terrorism is a very specific area of international and national concern and one that, in the UK, has a statutory definition in s.1 of Terrorism Act 2000. The concept of violence or violent crime is much broader. Criminalising the glorification of violence or violent crime has the potential to have a much wider reach, including material that some may view as mainstream. This could pose issues of proportionality.
5.111 The Association of Police and Crime Commissioners noted their concerns about the breadth of the term “glorification” in the context of violence and violent crime:452
We do have concerns about the broadness of the term “glorification” in this context, and regarding what harms such an offence would seek to redress that is not covered by other offences, e.g., those relating to incitement to violence. We are also concerned about how such a law could conflict with Article 10 Freedom of Expression in the European Convention on Human Rights.
5.112 The Bar Council of England and Wales argued that the existing offences under the Terrorism Act cater for “some types” of conduct glorifying violence or violent crime. They also noted the absence of evidence of need for a new specific offence:453
Existing anti-terrorism legislation (Terrorism Act 2006, ss.1 & 2 - encouragement of terrorism and dissemination of terrorist publications) caters for some types of misconduct involving glorification of violence or violent crime. We recognise, though, that the Terrorism Act 2006 does not cover all the forms of glorification cited by the Law Commission in this consultation. The most serious forms are covered by the Terrorism Act 2006. We think it important to point out that there does not appear to be an overwhelming evidential basis for creating an additional offence. In any event, the Bar Council is not in a position to inquire into, or provide, such evidence. We do register our concern as to the breadth of any additional offence.
5.113 A number of consultees, particularly legal consultees, noted the problematic nature of the glorification of terrorism offences. Further, they highlighted the risk that any new glorification of violence or violent crime offence may serve to alienate further already marginalised parts of the community in ways that is ultimately counterproductive.
5.114 Consultees that responded positively included those who actively deal with and combat violence and prejudice, for example Community Security Trust and the LGBT Fed. Their responses noted the prevalence of violence against members of their community, and the connection between online abuse/glorification and offline violence. These concerns are valid, and any connection between online “hate” or abuse and offline violence is worrying. However, most of these responses do not directly engage with the concerns we expressed in the consultation paper about expression and the potential for over-criminalisation, nor the ability of existing laws to grapple with these issues. We also believe, as we set out above and in the consultation paper, that the recommended harm-based offence will have some application in relation to harmful communications that may also glorify violence or violent crime.
5.115 Further, this is an area where a specific criminal response criminalising “glorification” may be difficult to target, and risks being a disproportionate response. Instead, the online harms framework and platform community guidelines may well be a more appropriate avenue. We note the detailed consideration given to this matter in the February 2021 report of the Commission for Countering Extremism.454 It is also important to note that communications containing threats of violence may also be addressed to a degree by our recommendations on threatening communications and our existing, yet to be implemented recommendations regarding threats and the OAPA 1861.455
5.116 A further refrain in consultee responses (again along the lines of concern we had already covered in the consultation paper) was that any offence of glorification of violence or violent crime would likely represent an unjustifiable, disproportionate interference with Article 10 and freedom of expression. A section of consultees noted the existing inchoate offences that may well be able to address culpable behaviour without having the “chilling” effect that any new glorification offence is likely to have.
5.117 Our concerns in the consultation paper about Drill music and “street illiteracy” are also reflected in a number of the responses from legal stakeholders. However, responses including those of English PEN and the Crown Prosecution Service echo our concerns and note that criminalisation would be a radical response to a problem that appears thus far to be poorly understood. As English PEN pithily put it, this provides an opportunity to contextualise rather than criminalise the behaviour.
5.118 After considering the detailed and thoughtful responses we received from consultees, we believe our provisional conclusion 456to have been correct: there is not sufficient justification for a new offence of glorification of violence or violent crime. There are two central aspects to this conclusion. First, the inherent vagueness in the concept of “glorification”, which we do not believe in this context would result in an appropriately confined offence. Secondly, the potential for a disproportionate and counterproductive impact on certain marginalised parts of the community.
5.119 Content relating to body modification overlaps with two other discrete topics in this report, encouragement or assistance of self-harm and glorification of violence or violent crime. As we discussed in the consultation paper, any broad offences of glorification of either self-harm or violent crime could be relevant to communications relating to body modification.457
5.120 However, as we discuss in detail in Part 2 above, we are of the view that there is insufficient justification for an offence of glorification of violence or violent crime. Further, as we set out in Chapter 7, our recommendations regarding encouragement or assistance of self-harm are tightly constrained to avoid their inadvertent application to vulnerable people who are not acting culpably.
5.121 The judgment of the Court of Appeal in R v BM 458confirmed that body modification procedures such as ear-removal, nipple-removal, and tongue-splitting can constitute the offence of causing grievous bodily harm with intent contrary to section 18 of the OAPA 1861, regardless of whether they were carried out with consent. This will be the case provided that the procedure in question does not fall within the “medical exception”, which would cover, for example, nipple-removal as part of a mastectomy.459 Following his unsuccessful appeal in R v BM, body modification practitioner Brendan McCarthy pleaded guilty to three counts of causing grievous bodily harm with intent and on 21 March 2019 was sentenced to 40 months imprisonment.460
5.122 As we set out in the consultation paper, the significance of the judgment in BM is that carrying out body modification procedures (provided they do not fall within the medical exception) can amount to a violent crime; therefore, communications promoting body modification would likely be covered by one or more of the existing inchoate offences of inciting, encouraging, aiding, or abetting violent crime.461
5.123 These offences are serious. For example, for the offences of encouraging or assisting an offence under sections 44, 45 and 46 of the SCA 2007, the maximum penalty is (as a general rule) the same as the maximum available on conviction for the relevant “anticipated or reference offence”.462 So, if a communication regarding body modification is found to amount to encouragement of causing grievous bodily harm with intent, this could, in theory, carry a maximum sentence of life imprisonment (the same as the maximum sentence for the reference offence, that is, causing grievous bodily harm with intent) - though we note that Brendan McCarthy’s sentence was significantly lower than this maximum.
5.124 This is our understanding of the position on communications relating to body modification under the existing law. It is not within our Terms of Reference to make proposals in relation to the existing law covering offences against the person (such as causing grievous bodily harm, with or without intent) or the existing inchoate offences.
5.125 We sought consultees’ views on the implications of possible offences of glorification of violence or violent crime, and glorification or encouragement of self-harm for body modification content:463
We welcome consultees’ views on the implications for body modification content of the possible offences of:
1) glorification of violence or violent crime; and
2) glorification or encouragement of self-harm.
5.126 The Justices’ Legal Advisers and Court Officers’ Service, formerly the Justices’ Clerks’ Society, agreed that where an existing offence is being encouraged, the inchoate offences that are already available can be used:464
We agree with your comments. We would only note that encouragement of criminal body modification is not an offence we see charged. However in an appropriate case, the existing offence of encouraging an offence is already available which renders specific legislation unnecessary.
5.127 ARTICLE 19 agreed with the cautious approach we took in the consultation paper: 465
As the Law Commission notes, it is entirely possible that some body modifications pertaining to one’s self-identity may inadvertently get caught in any new proposed offence of glorification of violence or encouragement of self-harm. We would strongly discourage the Law Commission from introducing any such new offence, that in our view would inevitably be unduly broad.
5.128 The Criminal Bar Association noted the problematic nature of “glorification” in criminalising violence or violent crime:466
We do not think the offence should be further modified. ‘Glorification’ is a problematic concept to tackle with the criminal law.
It is our view that any creation of offences of either glorification of violence or violent crime or glorification or encouragement of self-harm would necessarily include and thus criminalise body modification content especially in view of the case of R v BM. Therefore, unless there is a strong imperative and sufficient background evidence to justify a need to further amend the criminal law (terrorism matters aside, which are in any event outside the scope of this consultation), there should be no further offence created.
5.129 Many responses we received were from individuals who seemed unaware of the current legal position with respect to body modification. These responses stressed that if a person modifies their body of their own free will,467 or consents to another person modifying their body, that the law should not intervene. As we set out in the consultation paper and re-iterated above, the decision of R v BM 468confirmed that body modification procedures such as ear-removal, nipple-removal, and tonguesplitting can constitute the offence of causing grievous bodily harm with intent contrary to section 18 of the OAPA 1861, regardless of whether they were carried out with consent.
5.130 Other consultees noted the parallels between self-harm and body modification as they related to consent. As we set out in the consultation paper, the distinction between glorification and encouragement is worth noting. Our recommendation for an offence of encouragement of serious self-harm469 could potentially capture someone who posts material promoting self-administered body modification. It would only cover body modification that amounts to grievous bodily harm.
5.131 A number of consultees stressed that any new offence addressing “glorification” of self-harm was likely to be too broad. They noted that an offence predicated on “glorification” could bring non-culpable behaviour within its scope. We share these concerns and note that we make no recommendations for an offence of glorification of violence or violent crime. Regarding self-harm, our recommendations are carefully crafted to ensure they do not disproportionately apply to vulnerable people. They also only target encouragement and assistance of serious self-harm. This means many of the concerns raised by consultees about the potential over-reach of an offence predicated on “glorification” do not arise.
5.132 To the extent that a communication promotes self-administered body modification, this may be caught by a potential new offence of encouragement of serious self-harm, which we recommend in Chapter 7: . However, as we set out at length in the chapter, we are concerned that any such offence should not catch vulnerable people - such as people who are suffering from mental health conditions - who share content for the purposes of self-expression or seeking support. This concern also applies in the case of body modification.
5.133 The safeguards built in to the recommended offence of encouraging or assisting serious self harm include the fault element, requiring intention to encourage or assist self-harm to the level of grievous bodily harm; the high harm threshold of grievous bodily harm; and the requirement of (non-personal) DPP consent. These are all designed to ensure that vulnerable people are not inadvertently caught by our recommended offence of encouraging or assisting self-harm. Those safeguards will
also apply in respect of communications encouraging body modification. In light of the decision of R v BM128we believe this is justified.
6.1 In this Chapter, we recommend that a new offence be included in the Sexual Offences Act 2003 to address a harmful and increasing phenomenon, popularly - albeit, as we shall see, somewhat misleadingly - termed “cyberflashing”. Depending on the form of offending behaviour, this is a problem that is either not addressed well or not addressed at all by the current law.
6.2 The act of cyberflashing generally involves a person sending an unsolicited image or video recording of genitalia to another. It is important to distinguish cyberflashing from other forms of image-based sexual abuse where the victim is the subject of the image; in cyberflashing, the victim is not the subject but the recipient. The Law Commission has a separate project concerning the taking, making and sharing of intimate images of people without their consent.1 However, that is not the realm of cyberflashing.
6.3 In practice, recipients of cyberflashing will often not know the identity of the sender, and often the pictures or video recordings will be sent via a peer-to-peer protocol such as AirDrop, rather than via a telephone network or the internet - conferring on the recipient the twofold threat of a sender who is not only anonymous but also proximate. We have received significant evidence of the alarm and distress caused by this form of cyberflashing.
6.4 However, cyberflashing takes other, equally harmful forms. It is a particular problem on certain dating apps2 and social media, and can also take place between people who may know each other (through, say, their work). Recipients variously describe feelings not only of alarm and distress, but also of humiliation as a result of cyberflashing.
6.5 In this respect, cyberflashing therefore shares much with established forms of exposure (which is to say the act of a person exposing - or “flashing” - their genitals to another). This, when done intentionally and with the intention that someone sees their genitals and is caused alarm or distress, is criminalised under section 66 of the Sexual Offences Act 2003. However, there are two factors that limit the applicability of section 66 to cyberflashing. The first problem, as we noted in our consultation paper, is that it is not clear that section 66 would cover “non-live” instances of cyberflashing (ie photographs or recordings).3 Whilst some acts of exposure that take place online are live - such as an act of exposure over a video-calling facility4 - a significant
Taking, making and sharing intimate images without consent project webpage:
https://www.lawcom.qov.uk/project/takinq-makinq-and-sharinq-intimate-imaqes-without-consent/.
Mobile applications through which users can chat with other people, often (though not exclusively) with the intention of meeting or dating that person.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, para 6.116. See also Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, para 6.144.
See, for example, R v Alderton [2014] EWCA Crim 2204.
proportion of the offending behaviour constituting cyberflashing is not live, instead involving photographs or pre-recorded video.
6.6 The second limiting factor relates to the genitals exposed: specifically, whether they are the genitals of the person exposing them. That the exposed genitals are those of the exposer would seem to be the sine qua non of exposure; a person doesn’t “flash” using someone else’s genitals. In any case, it is a requirement of the section 66 offence, and not one liable to particularly tricky problems of proof. However, when images of genitals are sent to another, as in the case of cyberflashing, it may be neither obvious nor significant that the genitals are those of the sender. The harm experienced does not correlate with any sure knowledge that the image is that of the sender; indeed, in many if not most cases of cyberflashing, the recipient will have no idea whether the genitals in the image belong to the sender. Yet the harm is the same. Further, it may be difficult or impractical to require proof as to this matter. In this sense, cyberflashing differs from “flashing” in the traditional sense.
6.7 In our consultation paper, we proposed that the problem of cyberflashing be addressed by amending section 66 of the Sexual Offences Act 2003 to include the sending of images or video recordings of one’s genitals. This proposal addresses the first of the limitations we have identified here, but not the second. We were concerned to ensure that a law primarily focussed on exposure was not trying to cover too broad a range of behaviour, such as behaviour that may more accurately be described as sexual harassment or a communications offence.
6.8 However, we received a number of careful, well-evidenced, and rigorously reasoned responses to this part of our consultation paper. With the assistance of this evidence, we have had an opportunity to consider our position and have come to the view that there is a clear set of behaviour that can properly be addressed by an offence in addition to the existing section 66 offence, rather than by simply amending section 66 itself.
6.9 This additional offence is not an offence directed at the mere sending of nude images; it is not an offence directed at consensual sexting, or naturists, or consensual humour, none of which is the concern of a sexual offences regime. Many consultees were - in our view, rightly - concerned to note that any new offence should not be so broad as to criminalise consensual or non-harmful activity between adults. Instead, this is an offence directed at a form of behaviour that, like exposure, causes real harm and is accompanied by a clearly wrongful purpose.
6.10 This Chapter will proceed in two sections:
(1) the rationale for law reform; and
(2) the elements of the recommended offence.
6.11 In this section, we will address three matters that form the underlying rationale for recommending a new offence:
(1) the harm caused by cyberflashing;
(2) the prevalence of cyberflashing; and
(3) the inadequacy of the current law.
6.12 We did not ask consultation questions specifically directed at these matters, though, perhaps unsurprisingly, many consultees addressed these issues at various points in their responses. It is therefore worth noting the tenor of these responses, which confirms our account of these matters in the consultation paper (indeed, not a single response challenged our findings on these matters).
6.13 In our consultation paper, we noted some of the manifestations of cyberflashing and the attendant effect on victims.470 The feelings that were described in various testimony included shame, embarrassment, violation, fright, and vulnerability, to note but a few.
6.14 The responses that we received to our consultation have more than confirmed this to be the case. In an extensive joint response, Professor Clare McGlynn QC (Hon) and Dr Kelly Johnson noted a variety of harms resulting from cyberflashing:
Some victim-survivors have described their experiences of cyberflashing in terms of violation, describing how they felt: ‘utterly violated’; ‘really violated’; ‘incredibly violated’; ‘at its core, it’s very invasive’; ‘I just felt totally violated’. Marcotte et al (2020) found that almost one third of women reported feeling ‘violated’ after being sent unsolicited penis images...
Victim-survivors also report experiencing being embarrassed, disturbed, shocked, utterly horrified and ashamed, with one describing the ‘heatwave of embarrassment’ she felt:
‘The truth is, no matter how strong I thought I was, he turned me, with a picture, into a weak person, feeling humiliated and with no ability to stand up for myself . the incident still repeats in my mind’ (Boulos, 2019).
Combined, these experiences underpin a sense of humiliation, understood as infringing the dignity of the person (Gillespie 2019). The person is dishonoured and humiliated through a failure to show respect and through treatment of others as less than deserving of respect, and as means rather than ends.471
6.15 The Association of Police and Crime Commissioners referred to “the distress and feelings of violation that the sending of images or video recordings of the defendant’s genitals can cause.”472
6.16 Sophie Gallagher, a journalist who has worked extensively with victims of cyberflashing, noted that many victims “have changed their route home/train journey to avoid a potential repeat of the incident.”473
6.17 The Angelou Centre (a centre in the North East providing “support for black and minoritised women and children who have been subject to domestic and sexual violence”474) described cyberflashing as “a form of sexual violence and the impact of this on victim/survivors can be far reaching”.475
6.18 The National Police Chiefs’ Council noted that “victims feel traumatised as receiving [overtly sexual or abusive] messages can be unsettling, aggressive and intimidating.”476
6.19 Strikingly, the harm seems to be comparable with - or even worse than - traditional forms of exposure. According to the Suzy Lamplugh Trust, cyberflashing “can be equal to or more aggravating than offline sexual harassment”. 477Sophie Gallagher, in her response, remarked:
I interviewed many women over a period of years and for victims who have experienced both cyber flashing and traditional “anorak” in-person flashing - they say the impact was the same/if not worse for cyber flashing because of the anonymity given by the phone as opposed to being able to see who is flashing in person/other people around could also see and protect the victim...
... it is hard to see impacts of physical flashing as worse than digital flashing because it will not have a black and white scale for all victims. Some may find one format worse than the other because of specific triggers/touch points within their own experience.478
6.20 Professor McGlynn and Dr Johnson echoed these findings:
Women have frequently connected their experiences to physical sexual exposure: ‘it’s the same thing as flashing in public’. For many, the harm stems from the ‘well-founded fear’ of what might happen next, particularly in contexts where unsolicited penis images are sent in public from strangers. Women may not be harmed per se by being sent a penis image, but what it represents and what it might mean in practice; the implicit or explicit threat of further sexual assault.
Women recount, for example, feeling immediately ‘frightened’, ‘terrified’, ‘vulnerable’ and ‘exposed’ by acts of cyberflashing. They fear escalation of the actions, with women reporting feeling scared as to what might happen next. One victim-survivor stated: ‘with cyberflashing, because you don’t know who’s sent it, and you’re in a public space, that threat is never really eliminated’. Another said: ‘I was singled out, I was being targeted, and it felt very personal’.479
6.21 One thing that became particularly clear during our consultation was not just the harm that cyberflashing causes but also its prevalence. Our consultation paper included data from the British Transport Police showing a significant increase in reports of cyberflashing on public transport.480 Whether a result of rising cases or increased reporting, that data nonetheless suggested that cyberflashing was by no means an uncommon phenomenon. This was supported by the extensive testimonial evidence gathered by Sophie Gallagher, which we reference in our consultation paper.481 Responses to our consultation have further suggested that the scale of cyberflashing is significant and is certainly not diminishing.
6.22 In their response, the National Police Chiefs’ Council noted that “data shows that offences of cyber flashing are increasing at an alarming rate therefore we need to ensure legislation is robust enough.”482
6.23 Fix the Glitch, a UK-based educational charity raising awareness of online abuse, reflecting on the British Transport Police data, remarked that:
Organisations fighting sexual harassment and violence believe these numbers to be only the tip of the iceberg as many more cases remain unreported... Campaigners, journalists and activists have documented the concerning increase in cyber-flashing...483
6.24 In their joint response, Professor McGlynn and Dr Johnson referenced a wealth of data showing scale of the problem:
[A] YouGov survey in 2018 found that 41% of women had been sent an unsolicited penis picture; for younger women, this rose to almost half of women aged 18-24
(47%) (YouGov 2018). One study found 76% of girls aged 12-18 had been sent unsolicited nude images of boys or men (Ringrose, 2020).484
6.25 An important aspect of the responses we received was that they elucidated not only the scale of the problem but also the range of circumstances in which cyberflashing takes place. As we noted in our consultation paper, it is clear that cyberflashing in public places (such as public transport) via a peer-to-peer network such as AirDrop is one of the paradigmatic forms of offending, but Professor McGlynn and Dr Johnson noted that cyberflashing was prevalent across other media too:
Cyberflashing is experienced as routine by many using dating apps, including from strangers, acquaintances and potential daters. It is commonly experienced out of nowhere; other times following rejecting the man’s advances...
Cyberflashing is regularly experienced by women engaging in social media and other online technologies, in personal and professional capacities, by strangers, colleagues, acquaintances, family friends. It is also now taking place in online video conferencing with terms such as ‘zoomflashing’ and ‘zoombombing’ now into language reflecting a rise [in] various forms of online abuse, including online exposure and distribution of sexually explicit texts online.485
6.26 As we noted in our consultation paper, it is not clear that cyberflashing is addressed well or, in some cases, at all by the existing law. 486We were concerned that cyberflashing would not be covered by the existing exposure offence under section 66 of the Sexual Offences Act 2003, save if the exposure were “live” (through, say, a video-conferencing facility).
6.27 In the same way, the sending of the image being an essentially private act (in the sense of a communication passing between two people rather than an act witnessed publicly), it is difficult to see that the behaviour would constitute the common law offence of outraging public decency.487 This is not to say that the common law offence could not criminalise some forms of public exposure: were the sender to post an image over a public forum such as Twitter that is seen by all of their followers, that may well meet the conduct requirement of that offence. Were the image itself taken in a public place, then that act may meet the conduct requirement of the offence, though that is criminalising something other than the unsolicited sending of the image - the taking, rather than the sending.
6.28 Further, whilst the conduct may constitute an offence under section 127(1) of the Communications Act 2003 - being an indecent communication - the communication would have to be sent over a public electronic communications network to fall within the ambit of the offence. The requirement is satisfied if the image is sent over a mobile phone network or over the internet - both of which are, for the purposes of the offence, public electronic communications networks - but not if the image is sent over a private network or a peer-to-peer connection such as Bluetooth or AirDrop.
6.29 The offence in section 1 of the Malicious Communications Act 1988 is not restricted in this way: it is enough that indecent material is sent. It is worth recalling, though, that the prosecution must also prove that one of the sender’s purposes in sending the material was to cause distress or anxiety to the recipient.
6.30 In any case, we do not consider these two communications offences to be appropriate means to address cyberflashing. First, and perhaps most obviously, we are recommending that they be replaced with an offence based on likely harm. Second, even that offence would share with its predecessors one significant limitation: communications offences are not sexual offences, and so lack the procedural and evidential rules of a sexual offences regime, as well as the post-conviction ancillary orders. As we note in the following section, the overwhelming majority of consultees supported our proposal that cyberflashing be addressed through the sexual offences regime.
6.31 Our view that the current law was inadequate to the task of combatting cyberflashing was not challenged by any consultees. The Crown Prosecution Service (the “CPS”) explicitly agreed “that this area of law requires updating given the advances of technology”.488 This, coupled with the significant support for law reform, fortifies our view that change is necessary.
6.32 In this section, we recommend that a new offence of cyberflashing be included in the Sexual Offences Act 2003. In so doing, we will address three distinct questions:
(1) should the offence be classed as a sexual offence;
(2) what conduct should constitute the offence (the conduct element); and
(3) what should the defendant have known or intended, or had as their purpose, in order to commit the offence (the fault element)?
6.33 As we will note in the following discussion, there was very strong support for our proposal simply to amend the existing exposure offence in section 66 of the Sexual Offences Act 2003. This support was predicated on two contentions: (i) that the current law did not adequately address cyberflashing; and (ii) that, like exposure, cyberflashing was essentially a sexual offence.
6.34 However, that support for our proposal was heavily qualified; consultees were concerned that simply adopting the terms of the existing exposure offence - whilst an improvement on the current law - would still encounter real obstacles in addressing the full range of harmful behaviour. Though the reasoning differed across responses, these obstacles can be summarised generally as:
(1) the requirement that the sender send an image of their own genitals; and
(2) the narrow additional intent element of the original exposure offence (intention to cause alarm or distress).
6.35 We had anticipated these concerns and asked three consultation questions addressing these points specifically. Of course, good arguments can be marshalled to say that the identity of the genitals in question is irrelevant or, in establishing fault, that it should be enough that the sender lacked reasonable belief in consent. However, our concern was - and, as will be seen, remains - that accepting all of those arguments and formulating an offence on that basis results in an offence that is very broad. Such an offence would undoubtedly catch behaviours that are either not harmful or not criminally so. For example, a shared picture of a friend on a nudist beach may technically fall within the scope of such an offence due to the presence of genitals in an image but would not be meaningfully harmful. Similarly, other examples where an image of genitals is sent, but not coupled with sufficiently culpable intent such that the behaviour cannot be categorically characterised as criminally culpable, such as the genuine (even if misguided) attempt at humour, the immature juvenile.
6.36 Of course, we might say that these examples are some way from the paradigmatic case of cyberflashing and so unlikely to be prosecuted. However, whilst the CPS plays a vital role in the criminal justice system in prosecuting only those cases it is in the public interest to prosecute, it is not enough to rely on such prosecutorial discretion or guidelines to curb the reach of an offence drafted so broadly as to bring behaviour that should clearly not be criminal within its purview. It is also not enough to argue that innocuous messages, even if technically criminal, probably would not be reported to the police. The simple fact is that we do not know this to be the case. In any case, were such behaviour reported to the police, one could hardly blame them or the CPS for proceeding with a prosecution that quite clearly meets the criteria of the statutory offence.
6.37 Consultation question 24 asked:489
We provisionally propose that section 66 of the Sexual Offences Act 2003 should be amended to include explicitly the sending of images or video recordings of one’s genitals. Do consultees agree?
6.38 We addressed above the rationale for reform, including the inadequacy of the existing legal provisions and the harm caused by cyberflashing. Further, since the caveats that consultees provided in their response to this question fit within the other three questions that we asked, they are better addressed in the sections following this. The focus of the present discussion will therefore be on the third proposition underlying our proposal: that cyberflashing ought properly to be addressed through the sexual offences regime.
6.39 It is worth recalling what we said in our consultation paper on this matter:
.. .Cyber-flashing is, as a matter of common sense, conduct of a sexual nature. Further... those who have been subjected to cyber-flashing compare its impact to that of other sexual offences: for example, it can cause similar feelings of violation and sexual intrusion. Our proposed [communications] offence does not fully reflect this specifically sexual behaviour and harm. It covers a wide range of abusive communications, some of which may be sexual, and some of which are not. In our provisional view, there should be a clear option to prosecute cyber-flashing as a sexual offence in order to ensure that the nature of the offending conduct is more accurately labelled by the offence.
The second reason concerns the additional legal protections that apply in respect of sexual offences. Under section 103A of the SOA 2003, inserted by the Anti-social Behaviour, Crime and Policing Act 2014, these protections include Sexual Harm Prevention Orders: orders made by the court to protect the public or members of the public from the risk of sexual harm presented by a defendant. Under Section 103A, a court may make a Sexual Harm Prevention Order in respect of offences listed in Schedules 3 and 5, if it is satisfied that this is necessary for the purpose of:
(1) protecting the public or any particular members of the public from sexual harm from the defendant, or
(2) protecting children or vulnerable adults generally, or any particular children or vulnerable adults, from sexual harm from the defendant outside the United Kingdom.490
6.40 The overwhelming majority of consultees argued that cyberflashing should be an offence contained in the Sexual Offences Act 2003 - ie a sexual offence - albeit that many considered that it should be separate from the section 66 offence (a point we will address in detail below).
6.41 The Bar Council of England and Wales agreed with the provisional proposal:
We welcome the clarity that will be introduced with enactment of the newly proposed offence as outlined in Chapter Five [section 27(1)]. Alderton 491related to live exposure but the legal position at present is that it is unclear whether all forms of cyber flashing are contemplated by section 66 of the 2003 Act. We agree that this uncertainty is most directly and straightforwardly addressed by amending the statute.492
6.42 The CPS commented that it would:
...be helpful for cyber-flashing to be dealt with as a sexual offence, rather than under the proposed new harmful communications offence. This would facilitate correct labelling, appropriate sentencing and remedies such as Sexual Harm Prevention Orders. It could also facilitate automatic anonymity for victims if added to the Sexual Offences Act (SOA) 2003 and to section 2 of Sexual Offences (Amendment) Act 1992.493
6.43 The Association of Police and Crime Commissioners agreed with the proposals:
Given the distress and feelings of violation that the sending of images or video recordings of the defendant’s genitals can cause., we agree in principle that the Sexual Offences Act 2003 should be amended to include this, so that offences should carry with them the appropriate range of sentencing options (including Sexual Harm Prevention Orders).494
6.44 The Angelou Centre supported this provisional proposal:
This is a form of sexual violence and the impact of this on victim/survivors can be far reaching and thus the law should be reflective of this. In consultations completed with victim/survivors of racialised sexual harassment perpetrated online through explicit images, the impact can mirror that of violence already included within the Sexual Offences Act, primarily women feeling unsafe and violated. Many of the black and minoritised women and girls the Angelou Centre support who disclose experiencing sexual violence are blamed and viewed to have behaved shamefully/dishonourably and thus the receipt of sexualised images can lead to wider forms of violence, including domestic violence perpetrated within honourbased violence contexts. The law needs to reflect this level of risk.495
6.45 The Suzy Lamplugh Trust agreed with our proposal, acknowledging the importance of the option of Sexual Harm Prevention Orders in cases of sexual offending:
Suzy Lamplugh Trust agrees that section 66 of the Sexual Offences Act 2003 should be amended to include explicitly the sending of images or video recordings of one’s genitals. Considering the detrimental effects of cyber-flashing on the victim (and the fact that it can be equal to or more aggravating than offline sexual harassment), we believe it vital for the victims to have the choice 496to be protected by a Sexual Harm Prevention Order in such cases. By introducing this offence, the law would be able to send a clear message that online crime is as harmful and as punishable as offline crime.497
6.46 The Magistrates Association agreed, also noting the benefit of Sexual Harm Prevention Orders:
This seems a sensible proposal, especially as it can ensure Sexual Harm Prevention Orders can be used in response to offences.498
6.47 Professor McGlynn and Dr Johnson argued forcefully that any cyberflashing offence should be a sexual offence:
Singapore and a number of US states have recently adopted specific laws criminalising cyber-flashing and each are characterised as sexual offences. Scots law covers cyber-flashing as a sexual offence. Therefore, this approach follows international best practice...
Cyberflashing is a sexual intrusion which infringes victim-survivors’ rights to sexual autonomy and privacy. It is also experienced by some as a form of sexual assault. It is vital therefore that any new criminal law is framed as a sexual offence, ensuring appropriate recognition of the nature and harms of cyberflashing, and granting anonymity rights, special protections in court and suitable sentencing options. The specific cyberflashing laws that have been adopted in some US states and in Singapore have been enacted as sexual offences.499
6.48 Theirs was a position echoed in identical terms by the End Violence Against Women Coalition.500
6.49 Big Brother Watch, providing a general response to the Consultation Paper, argued that incidents of cyberflashing:
ought to be recognised as sexual offences and incur the sentencing and other consequences associated with sexual offences, given their specific and heightened risk.501
6.50 In contrast, The Free Speech Union disagreed, arguing that cyberflashing did not require a criminal response outside the communications offences:
We are not convinced of the need for this. If the general communications offence is limited to communications likely to cause harm (or as we would prefer intended to do so), we do not see why anything less, such as mere distress, should apply merely because the picture is of the sender’s or someone else’s genitals.502
6.51 While we understand the position put forward by the Free Speech Union, there are other areas of the law where the presence of genitalia changes the offence committed. Perhaps most obviously, a battery that involves genitalia is likely to constitute sexual assault.503 The responses that we have seen - from a wide range of different people and institutions, and many based on a wealth of first-hand testimony -make it clear that cyberflashing shares far more in common with sexual offences than it does with a communications offence.
6.52 The evidence we have received of the harm caused by cyberflashing, as well as support from significant bodies within the criminal justice system, has fortified our view that cyberflashing - just as with its “offline” equivalent, exposure - should be classed as a sexual offence. This would not only label the offence fairly, but would also allow for the range of evidential and ancillary orders that this would permit (both at investigation, trial and conviction stages).
6.53 For example, were cyberflashing listed in Schedule 3 of the Sexual Offences Act, then - pursuant to section 80 of the Sexual Offences Act 2003 - individuals convicted of cyberflashing could be subject to notification requirements (known colloquially as “being on the sex offenders’ register”). While notification requirements are not automatically imposed as a result of being convicted of any offence under the Sexual Offences Act 2003, many of the sexual offences in that Act are included in Schedule 3 and sufficiently serious convictions of those offences do lead to notification requirements being automatically imposed. Doing so for this offence would assist the police with management of offenders within the community. Notably, the section 66 exposure offence appears in Schedule 3, and we see no reason why cyberflashing ought not also to be included.
6.54 That this offence involves the sending of an image or video recording is uncontroversial. 504The image need not be sent electronically. There seems little reason to constrain the offence to an electronic communication; in our view, the offence should also cover situations where a perpetrator slips printed photographs of genitals into a victim’s bag. Indeed, as we noted in the consultation paper, one of the principal aims of these reforms has been to ensure that the law does not become redundant in the face of technological change: they are, in this sense, technologically neutral.505
6.55 Instead, if the paradigm behaviour with which we are concerned is the sending of an image of genitals by A to B, the more pertinent question is: whose genitals? Is it necessary that A sends a picture of their own genitals, or is it enough that they send a picture of anyone’s genitals?
6.56 Consultation question 25 asked:506
Assuming that section 66 of the Sexual Offences Act 2003 is amended to include explicitly the sending of images or video recordings of one’s genitals, should there be an additional cyber-flashing offence, where the conduct element includes sending images or video recordings of the genitals of another?
6.57 In our consultation paper, we noted our concern that, while the harm suffered by a recipient might be the same regardless of whether a person sent an image of his own genitals or those of another, it did not necessarily follow that the offence must bring both of these behaviours within its scope. As we said in the paper:
To do so would broaden the scope of the offence significantly and encompass behaviours that, in one sense at least, are quite different. For example, sending a publicly available image of a naked person (that includes their genitalia) to an acquaintance (who knows that the image is not of the sender) would seem to be a different order of threat from that posed by a stranger sending the same image, or where it wasn’t otherwise clear that the sender was not the person in the image. (For the avoidance of doubt, this is not to say that the former act would be harmless).
This would seem to suggest that the harm is not merely a function of whether it was the sender’s genitalia, but instead a more nuanced question of context and the apprehension of the recipient. Sending someone unwanted pornographic images may be harmful akin to forms of sexual harassment, but this is not to say that such behaviours should be governed by a law primarily focused on exposure.507
6.58 As will be seen in the responses below, one of the major obstacles inherent in an offence along the lines that we had proposed - and one that we had underestimated -is the problems of proof that might arise. Some consultees noted that it might be hard to prove the identity of the depicted genitalia in all cases. In particular, our attention was brought to the problem of “deepfakes” or otherwise altered imagery. Establishing whether these images were real or not could be prohibitively resource intensive.
6.59 Consultees also noted that, regardless of whose genitalia appeared in the picture, the harm was often identical. Indeed, many argued that it was irrelevant to the harm suffered. As we note below, we see merit in these submissions.
6.60 Refuge echoed both of the above concerns in their response:
...we are concerned that the proposals draw a distinction between cases of cyber flashing where the perpetrator sends images or videos of his own genitals and those where he sends images or videos of another person’s genitals. In terms of impact on the recipient of the images or videos, this distinction is unlikely to make a difference. Having separate offences could make the law challenging to use in practice and require investigation and evidence about whether the genitals in question were the perpetrator’s or those of another person, which seems to us an unnecessary evidential hurdle.508
6.61 Professor Alisdair Gillespie concurred:
I think cyber-flashing should be dealt with as an offence either under s.66 or as a stand-alone offence (s.66A?), but I don’t see the point in differentiating between one’s own and another’s genitals and I suspect it would be quite difficult to prove in courts509
6.62 The Association of Police and Crime Commissioners argued that:
From the perspective of the victim, if they receive an unsolicited image or video recording of a person’s genitals from a defendant, we believe that the impact on them would be the same, regardless of whether the genitals in question belong to the defendant or to a third party.
Additionally, defendants who share unsolicited images or videos of other people’s genitals are displaying similar - if not identical - sexually offensive behaviour, as defendants who have shared unsolicited images or videos of their own genitals.
Therefore, if the Sexual Offences Act 2003 is amended specifically to include the sending of images or videos of one’s own genitals, we would lean in principle towards suggesting that the sending of images or video recordings of other people’s genitals should be included in this. This would ensure that the law properly recognises the impact on victims, and that convicted offenders can be issued with Sexual Harm Prevention Orders, to prevent them from committing similar behaviour in future.510
6.63 The Magistrates Association agreed with our proposal though cautioned that relying on a harm-based offence may be afford insufficient protection:
We note the argument put forward that cyber-flashing which involved sending unsolicited or unwanted images or video recordings of the genitals of another would likely be covered by the harm-based offence proposed earlier in the consultation. However, we also note the arguments set out in respect of the previous question, explaining the importan[ce] of ensuring Sexual Harm Prevention Orders can be used in response to these types of offences. This suggests that it may be beneficial to have an additional cyber-flashing offence for this conduct that would not be covered by Section 66 of the Sexual Offences Act 2003.511
6.64 Dr Laura Thompson raised an interesting point concerning the problems of proof posed by developing technology, noting that the offence:
risks becoming quickly outdated (again) as technology evolves, specifically as advances in ‘deepfake’ technology make it easier to create images where it is almost impossible to tell whether they are real or faked.512
6.65 The End Violence Against Women Coalition disagreed with the proposed solution:
We reject the proposal to offer a two-part legal solution to cyber-flashing [as this] creates a hierarchy between different forms of cyber-flashing which is not justified on the evidence. There is no evidence that cyber-flashing is experienced as ‘worse’ or more harmful if it involves an image of the perpetrator’s own penis. In creating a hierarchy, there is a risk [that] not all cases of cyber-flashing will be taken seriously.513
6.66 Professor McGlynn and Dr Johnson also rejected the distinction that we had drawn, and argued that the structure of the proposed conduct element of the offence could be a result of misconceiving the offender’s motivations:
...the violation and intrusion, and possible fear and threat, experienced by the victim-survivor are not dependent on the knowledge that the penis in the unsolicited image is that of the perpetrator. To require proof that the image belongs to the perpetrator risks misunderstanding the nature of the experience and its attendant harms; it does not fully recognise the experience of victim-survivors.
It may be that proposals limiting a sexual offence to images of a perpetrator’s own penis assumes offenders are driven by motives similar to those of some physical ‘flashers’, namely the sexual dysfunction of exhibitionism, with often predatory consequences and a potential pre-cursor to other forms of sexual offending. Viewed from this perspective, the problem to be addressed in legislation is that of the individual exposing his own penis, with the fear of escalating sexual offending.
However, such an understanding neither captures the full range and extent of motivations of physical ‘flashing’, and even less so the wide-ranging purposes of cyberflashing, including inducing fear, alarm, humiliation and shame. It is vital that the multiplicity of motivations for cyberflashing are recognised and that the scope of any sexual offence is not unduly limited.514
6.67 As to the motivations of the perpetrators, Sophie Gallagher commented:
...in my case studies I have found less evidence that perpetrators are sending pictures/videos of other people (by this I mean that the pictures/videos look like home recordings/images/selfies as opposed to professional porn, for example).
When speaking to some men about why they do it the sexual satisfaction many get is often as a result of people seeing their genitals specifically, and the indirect humiliation that brings upon them, which they find gratifying. Although not all men do it for this reason - others do it for attention (any type), or to provoke reaction.515
6.68 Professor Tsachi Keren-Paz saw some benefit in using our proposed communications offence as a “fall-back” offence for images that were not those of the sender, but noted caution in drawing too great a normative distinction between the two cases:
.even if the analysis in the consultation paper is correct that sending an image of genitalia known to likely recipients not to be the sender’s [is] likely to cause less harm or threat (which is possible, but its normative implication might be overstated), an equal threat will exist if the likely recipient believes the genitalia is of the sender while it was not. The advantage of the proposed additional offence is that it could cover cases of victim’s mistaken belief that the genitalia was of the sender. Of course, this could be addressed by amending section 66 to include the sending of images or video recordings of or reasonably believed by the audience to be of one’s genitals.516
6.69 The Crown Prosecution Service, concerned with the problems of proof our proposed offence might create, argued:
Without the additional cyber flashing offence proposed above, it would be necessary for the prosecution to prove that the genitals shown are those of the suspect themselves unless the suspect has made an admission of this. There may be evidential challenges if the images do not contain the suspect's face (or some other distinctive feature which strongly links the image to the suspect) but depict the genitals only. This may in appropriate cases be overcome through the preferring of alternative charges if the suggested additional cyber flashing offence is enacted.517
6.70 Fix the Glitch’s submission reflected some of the concerns of other consultees as to the harm distinction:
Legislation which only criminalises the sharing or pictures of the perpetrator’s genitals fails to take into account the fact that such material - received without consent - causes distress regardless of whose genitals are depicted in the material. We therefore suggest that section 66 of the Sexual Offences Act 2003 be amended to include both the non-consensual sending of pictures or videos of genitals regardless of who the picture belongs to.518
6.71 Many consultees noted that women were most often victims of this behaviour; certainly, this is reflected in the data provided. However, one consultee submitted that this meant that the crime should be restricted to images of penises, and that our failure to discriminate in this way was itself an example of sexism.519
6.72 Taking the last point first, we are certainly aware that the majority of victims of this crime are women and that the majority of the perpetrators are men. This is a crime that clearly has a gendered element. Furthermore, we have seen evidence that minority groups are yet more susceptible to this type of behaviour. However, we have not been presented with sufficient evidence to conclude that this must exclude the possibility that men may be victims and women perpetrators. There has been no large-scale study. It would be striking were we to conclude that there were no circumstances in which the unsolicited receipt of an image of female genitalia could cause any of the harms described above. To the extent that comparisons with the offence of exposure are revealing, we know that men are victims of indecent exposure 520and it was noteworthy that Parliament, in passing the Sexual Offences Act 2003, specifically ruled out any amendment excluding female genitalia from scope:
...if the amendment were accepted, it would prevent the offence being used against females who expose themselves. While that is uncommon, as my noble friend said, it is not unheard of. The amendment would unduly limit the law. It would be saying that the man and woman who paraded naked together in public could not be convicted of the same offence. Given that the Government are anxious that the Bill should be non-discriminatory, it would be strange to accept an amendment that discriminates between men and women.
The primary concern should be the protection of the public.521
6.73 We concur: protection of the public is paramount, and failing to afford protection to a minority under the auspices of “labelling” is unjustified. We are not persuaded by the convoluted submission that our failure to discriminate is itself a form of sexism.
6.74 A slightly different but related matter concerns the extent to which women are forced or coerced to share images of themselves, and thus whether they may be unfairly caught within the scope of an offence that includes images of female genitalia. However, in such circumstances, the women would lack the mental element of the offence (which we address below). There is no need to restrict the applicability of the offence to images of male genitalia in order to circumvent this problem (to say nothing of those males who may be forced to share images of themselves).
6.75 As to the more general concerns, we are persuaded that, on balance, it would be better to have a single offence of cyberflashing - in addition to the existing section 66 exposure offence - that made no distinction between whether the depicted genitalia were those of the sender or otherwise. First, we agree that to require otherwise could present insurmountable problems of proof. We are not entirely persuaded that it would never be possible to establish this matter in court - the evidence on this point was rather general and non-specific - but we nonetheless take seriously the concerns expressed, especially by prosecuting authorities on this matter. All the more persuasive, however, was the observation that developments in technology, including Artificial Intelligence technology could make proof either impossible or so resource intensive as to be practically impossible. Even if we are not at this stage now, Dr Thompson was correct to note the potential for our proposed offence quickly to become outdated. This runs counter to our aim of technological neutrality, and desire to “future-proof” the offence as far as possible.
6.76 Second, we cannot ignore the wealth of evidence provided to us that the harm suffered was not altered in any way by knowledge or otherwise that the image was that of the sender. 522Part of our justification for proposing the offence be a sexual offence was rooted in the harm it caused; if the harm is the same in either case, it makes little sense to distinguish those cases as sexual and non-sexual.
6.77 This, of course, raises an interesting question that we alluded to in the consultation paper (cited above), namely what it is that specifically causes the harm: if not the act of “flashing”, is it really the fact that the image contains genitalia specifically that renders that image different from other sexual content? Why might we criminalise the unsolicited sending of one and not the unsolicited sending of the other? There may well be a very good answer to this, of course, but we did not consult on this question and so make no finding. For the purposes of a cyberflashing offence, it is enough to note that the evidence clearly shows that the paradigmatic and escalating behaviour involves the sending of images of genitalia.
6.78 We did consider whether the scope of the conduct element ought to be constrained by reference to the main subject matter (ie whether the image was substantively of genitalia). This might be one way to distinguish overtly sexual images from less sexual images. However, not only are we unconvinced this distinction really holds, but also there is no realistic way to draft such a distinction into law without either creating uncertainty or passing an easily circumvented law.
6.79 The effect of this will therefore be that the scope of the offence is broader than the offence we provisionally proposed in the consultation paper: sending an image or video recording that merely contains genitalia - anyone’s genitalia - would satisfy the conduct element of the offence. It is important to note, however, that the scope of the offence will be constrained by the fault element, to which we now turn.
6.80 The fault element of the offence concerns the defendant’s knowledge or intention: that is to say, what it is that they must know or intend in order to be guilty of the offence. The existing section 66 offence requires proof that the defendant intended to expose themselves, plus the additional intention that a person would see their act of exposure and be caused alarm or distress. Having proposed an amended section 66 as the vehicle for reform, proof of an intention to cause alarm or distress would have been required for the proposed cyberflashing offence. We were keen to know, however, whether there were other motivations or fault elements that we should consider. We also canvassed views on an alternative - whether awareness of a risk of causing harm should be sufficient.
6.81 We did consider in the consultation paper whether the offence should be based on knowledge of non-consent rather than a specific intention. 523However, we were concerned that an offence based purely on non-consent would bring within its scope too many different behaviours that were non-culpable, or at least not at a level sufficient to justify criminalisation.
6.82 One of the most useful results of the consultation was that it allowed us to understand more completely the range of harms and motivations attendant upon cyberflashing. The responses that we have received to these questions allowed us to reconsider the structure of the offence entirely. Two broad themes were apparent from the consultation responses: first, causing alarm or distress is too limited a range of motivations; second, that any offence must properly address the harm of invasion of a victim’s sexual autonomy.
6.83 As a result, rather than analyse each of the following two questions separately, greater clarity will be achieved by noting the tenor of responses to each question and then providing a fuller analysis at the end.
6.84 Consultation question 26(1) asked:524
Assuming that section 66 of the Sexual Offences Act 2003 is amended to include explicitly the intentional sending of images or video recordings of one’s genitals, should there be an additional cyber-flashing offence, where a mental or fault element includes other intended consequences or motivations, beyond causing alarm or distress?
6.85 #NotYourPorn submitted that, whilst alarm or distress was too limited a test, the test should be based on non-consent as a matter of fact:
This is creating an unjustified hierarchy of cyber flashing offences and [the] mental element threshold is too limited. Causing alarm or distress does not take into account the reality of the situation - the evidence shows that men cyber flash victims for a host of other reasons such as humiliation. In any case, the law should focus on non-consent not the motivation of the perpetrator.525
6.86 Professor McGlynn and Dr Johnson argued that motivation was of subordinate relevance to non-consent:
Cyberflashing is problematic because it is non-consensual conduct of a sexual nature. Distributing penis images is not per se wrongful, but doing so without the consent of the recipient is. The non-consensual act breaches the individual’s rights to sexual autonomy, regardless of the motive of the perpetrator. A focus on nonconsent as the core wrong is the approach of US states which have adopted specific cyberflashing offences.
While there are real challenges with proving consent in sexual offence cases, a major impetus for a cyberflashing law is to raise awareness, challenge the normalisation of the practice, aid prevention and education initiatives and to let victims know that their experiences are understood and recognised. These aims are met by focussing on the core wrong of non-consent and therefore justify this focus, in preference to a law requiring proof of specific motives.
Motive requirements (such as in the laws on image-based sexual abuse) invariably mean that only some forms of abuse are covered and create a hierarchy of abuses which does not reflect victim’s experiences. For example, a law requiring proof of intention to harm (or awareness of risk) will likely exclude some forms of cyberflashing, as well as making prosecutions more difficult because of the threshold being introduced.526
6.87 The Angelou Centre argued that the offence should be based primarily on nonconsent but also noted that, in respect of motivations:
by broadening the offence to cover other intended consequences it allows the prosecution of perpetrators that have committed the offence not to cause distress or alarm but for example, for sexual gratification, power or to humiliate the victim/survivor - which is often present in the perpetration of sexual violence. For example, the use of humiliation as a motivation is used in the Upskirting Legislation. This is particularly relevant for black and minoritised women and girls, who have disclosed that sexualised images have been sent by perpetrators with the intention of undermining their ‘honour’ within their families and wider communities - by the law encompassing other intended consequences or motivations, the lived of experiences of black and minoritised victim/survivors can be reflected in the law.527
6.88 The End Violence Against Women Coalition, arguing that alarm or distress was too narrow a range of intentions, recommended that:
...any motivations required to be proven be extended to include humiliation, as discussed in previous questions, which is used in similar legislation, including in Scotland and Singapore and in the English ‘upskirting’ legislation.528
6.89 The Crown Prosecution Service noted that the proposed offence did not address sexual motivations:
As set out in our response to Question 24, we agree that the offence should include other intended consequences or motivation beyond causing alarm or distress to cover the example provided in paragraph 6.117 of the Consultation Paper where the defendant saw their behaviour as a sexual advance rather than to cause ‘alarm or distress’. In the anecdotal experience of CPS prosecutors, defendants expose themselves for a range of reasons including for the purpose of sexual gratification and ‘having a laugh’.529
6.90 The Association of Police and Crime Commissioners echoed the concern that alarm or distress was too limited:
In principle, we believe that if a mental or fault element were to be included in the amends to section 66 of the Sexual Offences Act 2003, then it should go beyond an intention to cause alarm or distress: regardless of what the defendant’s intention for exposing an image or recording of their genitals may have been, this is unrelated to the potential impact that may be felt by the victim.530
6.91 Dr Laura Thompson submitted that:
The mental [or] fault element should include sexual gratification as a motive (as per the Singaporean sexual exposure law), as well as intent to cause humiliation (as per the current upskirting law). But this should be an amendment to section 66, not an additional offence.531
6.92 The Criminal Bar Association remarked:
...given the wide ranging circumstances of the sending and receipt of such images in this digital age we think for any cyber-flashing offence, the mental element of the offence could / should be expanded beyond a defendant intending to cause alarm or distress.532
6.93 Other consultees were less convinced of the need for expanding the range of culpable motivations. The Bar Council submitted that:
A person’s motive for cyber-flashing should be distinguished from their intent. It is difficult to conceive of circumstances of cyber-flashing, other than to an intimate, where the intention is other than to cause at the bare minimum alarm or distress. If it was a communication to an intimate, then, as suggested in the consultation, it would seem not to cross the threshold of criminality, even in the absence of express consent.533
6.94 Professor Alisdair Gillespie argued that “alarm or distress is sufficient.”534 Fair Cop - a free-speech group - also argued that “alarm or distress is surely sufficient.”535
6.95 Kingsley Napley were especially concerned about the potential scope of a broadened offence:
The legislation must be tightly worded to ensure that the criminal behaviour captured reflects the actions of individuals who have engaged in obviously culpable behaviour and not those acting in a careless, non-criminal way.
Identical behaviour can carry varying levels of culpability in this context depending on the circumstances in which communications are passed. Reasonable behaviour for individuals involved in a relationship for example, would be different to behaviour expected between two strangers...
There is a real risk that the definition of the offence could be broadened without thorough consideration being given to the repercussions, which could lead to the criminalisation of poorly judged, innocently sent communications.536
6.96 Kingsley Napley also noted (in response to Question 25) the potential effect on young people that could result from an overly broad mental element:
Consideration must be given in relation to the intention of the act, particularly where the elements of the offence are committed by children.
Criminalising acts which were intended to be non-malicious, albeit distasteful jokes would hugely broaden the criminalised group and would likely result in convicting individuals who have acted in a certain way due to immaturity and poor judgement, rather than with criminal intent.537
6.97 Though not directly related to the consultation question, and for want of a more appropriate place, it is also worth noting here Kingsley Napley’s comments on sentencing:
Different maximum sentences between children and adults should be considered to provide proportionate consequences to the careless rather than criminal acts of children.
Sentencing differentiation is employed elsewhere in relation to similar offences committed by children, for example section 13 of the Sexual Offences Act 2003 provides for a different maximum sentence for under 18s who have committed various sexual offences set out in sections 9 to 12 of the act.538
6.98 Consultation question 26(2) asked:539
Further, should the defendant’s awareness of the risk of causing harm (whether alarm or distress, or otherwise) be sufficient to establish this mental or fault element of the cyber-flashing offence?
6.99 The Justices’ Legal Advisers and Court Officers’ Service (formerly the Justices’ Clerks’ Society) agreed and noted:
Consent is a part of it in our view, but not all of it as you say. Perhaps the new offence could include “without consent”, but also “with intent to cause alarm or distress or awareness of a risk of doing so”.540
6.100 Alisdair Gillespie agreed, commenting that “...awareness of risk should suffice.”541
6.101 The Angelou Centre supported this approach:
This would allow for the perpetrator to be prosecuted when their main intention is not to cause alarm, distress or any other required motivation, but they were aware that their behaviour would likely cause alarm or distress. It also continues to ensure that the focus is on the behaviour of the perpetrator, rather than the victim/survivor and how she/he reacted or did not react.542
6.102 The Association of Police and Crime Commissioners agreed:
We think that the defendant’s awareness of the risk of causing harm should be sufficient to establish the mental or fault element of the cyber-flashing offence; this will ensure that consensual sharing of images or recordings of genitals between adults is not unduly criminalised.543
6.103 The Magistrates Association agreed and added the following:
This seems a sensible approach to be taken in respect of the mental element of a cyber flashing offence. The discussion in the paper around whether express consent should be required to receive such an image is an interesting one. We agree that you would not want to criminalise actions between partners or friends, where there is an expectation based on their relationship that such an image would be welcomed.544
6.104 Fair Cop disagreed: “We think that it is important to show intent to cause harm if the criminal law is to get involved.”545
6.105 The Criminal Law Solicitors’ Association also disagreed, commenting that this approach was “far too subjective.”546
6.106 The Criminal Bar Association were also concerned with this suggestion:
This is an area fraught with potential difficulties. As an example, the extension of this offence as suggested may lead to the increased criminalisation of young people who may misjudge their actions and their audience. The consequences of doing so may be severe and disproportionate. In particular, the automatic placement on the sex [offenders’] register may hold back, divert or marginalise young people who may in the particular circumstances have acted without malice and having caused relatively little or no harm. Neither should adults be faced with criminality when they intend no harm but merely misjudge at the relevant moment of sending an[] image believing that it may be well received. In order to temper this we feel that where the defendant’s awareness of the risk of causing harm (whether alarm or distress, or otherwise) is deemed sufficient to establish the mental element, the court should also have to take into account among other things, the context of the sending and any specific features of the defendant (and recipient).547
6.107 Gina Miller, echoing the concerns of Kingsley Napley and the Criminal Bar Association, asked “Might it be appropriate to include an age limit to this offence?”548
6.108 We have real sympathy with the argument that an absence of consent is common to all acts of cyberflashing (ie those communicated images we would want to criminalise). However, we are not persuaded that the touchstone of criminal wrongfulness lies in the absence of consent alone. The threshold of criminality for this sort of conduct must be higher than that. Clearly, some acts are so manifestly wrongful that non-consent alone should be enough to justify criminalisation. For example, in our consultation paper on Intimate Image Abuse, we argue that the violation involved in taking or sharing an intimate image of someone without their consent is “an attack on something fundamental and precious to the victim’s personhood”.549 It is an act that so violates a person’s sexual autonomy, bodily privacy and dignity that no additional intent element (such as intention to humiliate etc) should be required. However, while some of these matters are clearly implicated in cyberflashing - sexual autonomy would appear to be an obvious example - it is not clear that all are (the victim’s bodily privacy, for example). In any case, the infringement of sexual autonomy that results from cyberflashing - which we do not doubt - would seem to be a different order of harm than the violation of sexual autonomy that is the non-consensual taking or sharing of an intimate image.
6.109 While we appreciate that the harm experienced by victims is a product (in part) of the violation of their autonomy, we do not believe that a non-consent approach to criminal liability is the best way to criminalise cyberflashing.
6.110 Whilst a lack of consent and corresponding lack of belief in consent may be common to all forms of culpable cyberflashing, it does not follow that they are unique to culpable cyberflashing. If the recipient does not know that the sender is about to send a message, the question of whether the recipient consents is something of a fiction: of course there is no consent, because the question of consent (absent prior discussion) has not arisen. This is true of any situation where someone receives an unexpected message: the recipient has not had the opportunity to put their mind to the question of consent. Therefore, to ask whether the recipient “consented to” the message risks becoming a determination of whether or not the message happened to be welcome -but this is a determination after the fact; the crime is already complete at this point. The question of whether the message happened to be welcome or not tells us little about whether there was consent at the point the message was sent.
6.111 The reverse is also true. The question of whether there was consent is not sufficiently predictive of harm as to define the set of criminally culpable acts. If it is true that any message sent without explicit prior consent is strictly non-consensual - and this is surely true by definitional fiat - the effect of criminalising based on lack of consent would criminalise any image sent without explicit prior consent. As we note below, such a set would encompass many acts that ought not properly to be described as criminally wrongful acts.
6.112 We share the concerns of a number of consultees that there is a real risk of criminalising those who might misjudge a situation - whether that be, for example, misjudged humour or a misjudged attempt at intimacy - and that this risk falls especially hard on the young (including young adults). Given that the image need only contain genitalia, it would be enough to establish criminal guilt that someone sent to a friend a picture of themselves fully clothed on a busy nudist beach and did so without the friend’s consent. One can think of numerous other examples along these lines -from the prosaic to the distasteful to the mildly odd - where it is not obvious that the threshold of criminality is met, to say nothing of the fact that they fall a long way outside the paradigmatic case of cyberflashing.
6.113 As we note at the beginning of this chapter, it is no answer to this problem simply to say that the police would never investigate or the CPS would never prosecute. It is of course true to observe that the CPS has prosecutorial discretion in respect of all offences that are referred to it. No law can ever meet precisely the boundaries of the behaviour we might wish to criminalise; wrongfulness is a continuum whereas the criminal law is binary - either conduct falls on one side of the line or the other. This is an oversimplification, of course, but the point is that inevitably some overcriminalisation can happen. When it does, it can be dealt with using prosecutorial guidelines, for example, or by applying the public interest test, and so on. We return to this point below. Our concern here is that the balance is wrong: we would be proposing an offence that relies too heavily on its not being prosecuted: hoping that a loving partner will not be reported, hoping that the CPS won’t prosecute a misguided young person, hoping that a court would not convict a vulnerable person.
6.114 It is for the same reason that we consider “awareness of a risk of harm” to be too low a bar for what is - being a sexual offence - a relatively serious offence, and certainly one that carries weighty moral and social opprobrium. The sender would not need to anticipate likely harm, only to appreciate that there was a risk of that harm - whether or not they believed on balance that their message would be well received. We do not consider that such scenarios are for the criminal law to address.
6.115 Although we were not persuaded to change the basis of fault from intention (save for the addition of a purpose, which we will address in due course), we were persuaded that intention to cause alarm or distress was too narrow a fault element, reflecting neither the harm suffered by many recipients nor the intentions of the senders. Many consultees noted that humiliation was regularly a result of and clearly intended in many instances of cyberflashing, and rightly noted its inclusion in the “upskirting” offence found in section 67A of the Sexual Offences Act 2003.
6.116 Further, a number of consultees noted that obtaining sexual gratification was frequently one of the purposes underlying cyberflashing. While we have no reason to doubt this - indeed, it almost seems self-evidently true - it cannot just be tagged on to the list of “intentions”. Obtaining sexual gratification differs from the malicious intentions (causing alarm, distress, and humiliation) in this context because sending a person an image of genitals for one’s own sexual gratification is, on its own, not wrong. Indeed, it may be welcome. The harmful outcome is not embedded within the intent. However, in the right circumstances, an offence that criminalises cyberflashing done for a sexual purpose can better recognise the harm inflicted by the invasion of a victim’s autonomy. For this reason, we believe a different fault element is appropriate in certain circumstances where the defendant is acting for the purpose of obtaining sexual gratification.550
6.117 In our view it is appropriate to criminalise cyberflashing where an individual sends an image for a sexual purpose, reckless as to an adverse, harmful consequence (ie distress, alarm or humiliation). Recklessness requires proof of an awareness of the risk of a result coupled with the risk being unreasonable to take.551 Importantly, this would cover the paradigmatic cases where a stranger on public transport sends a relevant image; few adults would be unaware of the risk of harmful consequences when sending genital images to strangers, and it would seem highly unlikely that a defendant could run successfully an argument that it was nonetheless reasonable to take such a risk. It would likely also avoid criminalising those instances where someone sent a message uncertain of whether there was consent but where they genuinely believed that no harm would result (such as a loving relationship) or where, through lack of maturity, they were entirely unaware of such a risk (such as, perhaps, with youths).
6.118 We believe that this approach better reflects the culpability of the defendant at the point the message was sent (ie when the offence is complete). Further, it addresses the risks of over-criminalisation that attend an approach based on non-consent. It is also appropriate to note that while we do not believe that a non-consent approach is the best way to criminalise cyberflashing, non-criminal law responses to potentially harmful - albeit not criminally culpable - conduct have an important role to play.552
6.119 While the above, in our view, ensures that the offence will only capture culpable behaviour, we acknowledge that some extra care may be needed with potential youth offenders. Prosecutorial guidance can still play a valuable role. To that end, we believe that specific prosecutorial guidance should be formulated to address cyberflashing similar to that addressing section 67A of the Sexual Offences Act 2003, which contains the upskirting offence.553
6.120 What we therefore recommend is an offence of cyberflashing that has two alternative fault elements:
(1) if the sender intends to cause alarm, distress or humiliation; or
(2) if the sender’s purpose in sending the image or video recording is to obtain sexual gratification and the sender is reckless as to whether alarm, distress or humiliation are caused.
6.121 The evidence provided by consultees of the scale of cyberflashing and the harm it causes was striking. Although the data is not comprehensive, it is clear that consultees with extensive experience in the field - whether through criminal justice, academia, support networks or otherwise - are alarmed at the problem. There is no doubt that the current law is inadequate to the task of protecting people from cyberflashing.
6.122 However, as a number of consultees submitted, there is real risk in over-criminalising in this field. We are recommending a sexual offence to tackle an undoubtedly sexual behaviour, but we must be careful not to sweep up alongside the malicious the merely misguided, to gather up the vulnerable with the vicious.
6.123 It is our view that the problem of cyberflashing is best addressed by a specific offence of cyberflashing, ie an offence in addition to the exposure offence found in section 66 of the Sexual Offences Act 2003. This offence ought to be a sexual offence within that Act.
6.124 The offence we recommend would be technologically neutral; it does not matter over or through what medium you send the image, nor should the offence be constrained to images of the sender’s genitals. It is perhaps interesting to note in passing that the conduct element of the “cyberflashing” offence thus involves nothing inherently related to computer networks nor anything inherently related to public exposure.
6.125 However, this broad conduct element is balanced by a fault element that focusses on specifically wrongful behaviour, whilst at the same time having greater scope than exists under the current exposure offence.
6.126 Given the range of harms that have been described to us that result from cyberflashing and the attendant motivations, it seems that there may be reason to revisit the relatively restricted fault element of the original exposure offence (though, given that we have proposed a separate cyberflashing offence, that now lies outside the scope of this Report).
6.127 Finally we also note that in our view, the harm as described to us by stakeholders, and the conduct and fault elements as we outline above, seem to align with the offence having the same maximum penalty as the existing offence in section 66 of the Sexual Offences Act 2003. That offence is an “either-way” offence, liable to up to 6 months’ imprisonment on summary conviction or two years’ imprisonment on conviction on indictment.
6.128 Finally, part of the reasoning underlying our proposal that cyberflashing be considered a sexual offence were the ancillary orders that would be available were it included in Schedule 3 of the Sexual Offences Act 2003. As we noted above, consultees were overwhelmingly in support of this, citing the availability of Sexual Harm Prevention Orders and notification specifically. As our proposal was to amend the existing section 66 offence, which carries with it existing triggers under Schedule 3, we did not consult on what specific elements of a new offence should trigger an SHPO or notification requirement. However, there would seem to be no compelling reason for excluding a cyberflashing offence from the auspices of Schedule 3 of the SOA 2003 with a threshold similar to that applicable to exposure.554 The two separate fault elements in our recommended offence are similar to the offence in section 67A of the Sexual Offences Act 2003, which only triggers notification requirements under Schedule 3 when the defendant’s purpose was obtaining sexual gratification. 555However, in our view, the recommended offence is more similar in both substance and form to the offence of exposure in section 66 of the Sexual Offences Act 2003. As such, we think it more appropriate that notification requirements mirror those for the offence of exposure.
6.129 We also consider that, in line with other sexual offences, automatic lifetime anonymity and special measures for victims 556should be available in respect of offences of cyberflashing. These issues were considered in detail in our project on intimate image abuse.557
6.130 Objections can be made to imposing restrictions on the reporting of a complainant’s identity. The principle of open justice is fundamental and is not to be departed from without good reason. One practical concern is that in the case of a false allegation, a defendant may lose the opportunity to trace witnesses who could provide important exculpatory evidence regarding the complainant relevant to their case. More commonly it is argued that without anonymity for the defendant there is no parity of treatment and that this conflicts with the principles of open justice and fairness. However, automatic anonymity in relation to sexual offending has its origins in the 1975 Report of the Helibron Committee, 558and a range of statutory exceptions to the principle of open justice make specific provisions for anonymity. 559The question, then, is not whether anonymity can ever be justified but whether the public interest requires automatic (as opposed to discretionary) anonymity to be extended to victims of cyberflashing.
6.131 As we set out in detail above, in our view it is essential that cyberflashing be classified as a sexual offence in the SOA 2003. The other offences in that Act attract automatic lifetime anonymity, including the “in-person” exposure offence in section 66. 560In our view, not to extend automatic lifetime anonymity to an offence of cyberflashing would create unacceptable confusion and inconsistency. Automatic anonymity would ensure that victims of cyberflashing feel confident to come forward and report the offending, and that they do not see any lack of anonymity as a “barrier” to reporting.
6.132 Further, as a sexual offence, it is our view that complainants in cyberflashing matters ought to be automatically eligible for special measures at trial unless witnesses inform the court that they do not wish to be eligible. 561Another important safeguard for victims of cyberflashing at trial is the restrictions on cross examination that apply in trials involving sexual offences. 562Individually and in combination, we believe these measures will go some way to ensuring victims’ confidence in reporting cyberflashing, as well as limiting distress experienced by complainants in the trial process. This would similarly mirror the protections currently afforded to complainants of offences of “in-person” exposure.563
6.133 We recommend that the Sexual Offences Act 2003 be amended to include an offence of cyberflashing with the following elements:
(1) The defendant (A) intentionally sent an image or video recording of any person’s genitals to another person (B), and
(2) either-
(a) A intended that B would see the image or video recording and be caused alarm, distress or humiliation, or
(b) A sent the image or video recording for the purpose of obtaining sexual gratification and was reckless as to whether B would be caused alarm, distress or humiliation.
6.134 We recommend that Schedule 3 of the Sexual Offences Act 2003 be amended to include the offence of cyberflashing so that notification requirements under Part 2 of the Sexual Offences Act 2003 are imposed when an appropriate seriousness threshold is met.
6.135 We recommend that Sexual Harm Prevention Orders be available for all cyberflashing offences.
6.136 We recommend that victims of the new cyberflashing offence should have automatic lifetime anonymity.
6.137 We recommend that victims of the new cyberflashing offence should automatically be eligible for special measures at trial.
6.138 We recommend that restrictions on the cross-examination of victims of sexual offences should extend to victims of the new cyberflashing offence.
7.1 Encouragement of suicide or self-harm falling short of suicide is of great concern both
online and offline. While we raised concerns about how the criminal law might best tackle encouragement of self-harm in our consultation paper, we received detailed, well-evidenced responses from consultees that have been of great assistance in considering the best way forward. While a specific offence addresses the equivalent behaviour in the context of suicide, there is currently no offence that adequately addresses the encouragement of serious self-harm.
7.2 Unlike many other areas of criminal law that target “encouragement”, there is no underlying statutory offence of harming oneself, nor are we considering such an offence as the criminal law is not the right response to such a complex and sensitive issue. 564Any criminal law solution must be properly constrained to ensure that it does not disproportionately impact vulnerable people who harm themselves.
7.3 The issue of glorification or encouragement of self-harm was addressed in our scoping report 565and in Chapter 6 of the consultation paper. We asked two specific questions and received a diverse range of stakeholder feedback which is analysed below.
7.4 Part 1 of this chapter sets out the existing legal position and our position as put forward in the consultation paper. An analysis of the issues that arose in consultation follows in Part 2. Part 3 develops a proposal to address the issue and ultimately contains our recommendations for reform.
7.5 As we set out in the consultation paper, self-harm content online is a worrying phenomenon. We raised concerns about the vulnerability of those who post self-harm content and noted the importance of ensuring that mental health solutions are central to any attempt to address the matter. As explored below, after reflecting on consultees' views and considering the various potential options for criminal regulation, we believe a narrow offence with a robust fault element that targets deliberate encouragement of serious self-harm is appropriate.
7.6 Ultimately, we recommend that Government enacts a targeted offence to address the matter.
7.7 We considered the various harms relating to the glorification or encouragement of self-harm in the consultation paper. Aspects of this behaviour were not adequately addressed by the existing communications offences in the Communications Act 2003 (“CA 2003”) and Malicious Communications Act 1988 (“MCA 1988”):3
3.153 The existence of online content glorifying, encouraging, or promoting selfharm and suicide has attracted significant media attention and has been linked with the deaths of children and young people. For example, the so-called “Blue Whale Challenge” - an online “suicide game” which sets daily “challenges” for “players” -has been well-documented. Daily “challenges” start with, for example, “wake up in the middle of the night", then escalate to “cut a blue whale into your arm”, and finally, to suicide.4 This is an extreme example of content promoting self-harm. At the more insidious end of the scale are websites and social media pages promoting strict diets that may amount to eating disorders or “orthorexia”.5
3.154 While encouraging or assisting suicide is a specific offence, criminalised under the Suicide Act 1961, encouraging or assisting self-harm is not. As we note in the Scoping Report, there is an argument that glorifying self-harm may be an inchoate offence.6 We discuss this in detail in Chapter 6. Here, we simply note that, unless a communication glorifying, encouraging, or promoting self-harm crosses the threshold of “obscene, indecent, or grossly offensive”, it cannot be prosecuted under section 127 CA 2003. For similar reasons, it may not be caught by section 1 MCA 1988. It is, therefore, another example of potentially harmful communication that is arguably under-criminalised by the existing law.
7.8 The behaviour and impacts of glorification or encouragement of self-harm were considered in detail in the consultation paper.7 We noted that the literature tends to refer to “non-suicidal self-injury” (“NSSI”) content, which may include a range of conduct including (but not limited to):
- cutting (which seems to be particularly prevalent),
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, 3.153-3.154. Citations in original. The issues surrounding the “Blue Whale Challenge” and other “digital ghost stories” are explored in more detail in Part 2 below.
See, for example, A Adeane, Blue Whale: What is the truth behind an online 'suicide challenge'? (13 January 2019), available at: https://www.bbc.co.uk/news/blogs-trending-46505722 (last visited 13 July 2021).
See, for example, S Marsh, Instagram urged to crack down on eating disorder images (08 February 2019), available at: https://www.theguardian.com/technology/2019/feb/08/instagram-urged-to-crack-down-on-eating-disorder-images (last visited 13 July 2021). As defined in the Oxford English Dictionary (2021), “orthorexia” means: an excessive concern with consuming a diet considered to be correct in some respect, often involving the elimination of foods or food groups supposed to be harmful to health.
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, para 12.94.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.177 to 6.182.
- disordered eating,
- bruising,
- scratching, or
- substance abuse. 566
7.9 We outlined the vagueness of the term “glorification” in the context of NSSI, and noted the tension between content representing NSSI without endorsement and the statutory definition employed in the context of terrorism offences only requiring “any form of praise or celebration.”567 This tension was raised with us by consultees. We are conscious that, given the nature of NSSI content, and those who post it, an offence that does not have an appropriately narrow scope risks impacting many who are not deliberately or maliciously encouraging self-harm. This is explored in greater detail in Part 3 below.
7.10 The literature discusses the respective benefits and risks of NSSI content. While the benefits include mitigation of isolation, encouragement of recovery and curbing of NSSI urges, the risks include reinforcement, triggering and stigmatization of NSSI.568 As we discussed in the consultation paper, positive portrayals of NSSI have been found to be “few and far between” and sharing of NSSI content tends not to be maliciously motivated but instead to communicate personal experiences of self-harm.569
7.11 This is not to say that there is no malicious NSSI content. As we considered in the consultation paper, and as relayed to us in our consultation meetings and consultation responses (considered in detail in Part 2 below), there are also examples of deliberate and malicious encouragement of self-harm.570
7.12 Two National Crime Agency (NCA) cases were brought to our attention during the consultation. Each provides an illustration of the type of conduct resulting in self-harm we believe should be criminal, involving exploitation of vulnerable people. It appears that in some circumstances severe self-harm can occur through incitement or coercion of vulnerable people in the context of other offending, including sexual offending and blackmail. Whether this conduct fits the scope of an offence of “encouragement” is considered further below:
(1) R v Falder 571is a judgment of the Court of Appeal (Criminal Division). The
defendant pleaded guilty to a total of 137 offences all either of a sexual nature or sexually motivated. The offences related to his manipulation of numerous victims over the internet. It is also worth noting that his guilty pleas included numerous offences under the Malicious Communications Act 1988 (though given the severity of his conduct, what was caught under the MCA would likely fall under our provisionally proposed “harm-based” offence). Falder’s offending can be summarised as follows:572
Falder was a university academic who is now serving a sentence of 25 years’ imprisonment after admitting 137 charges. His conviction followed an investigation by the NCA into horrific online offending which included encouraging the rape of a four year old boy. Falder approached more than 300 people worldwide and would trick vulnerable victims - from young teenagers to adults - into sending him naked or partially-clothed images of themselves. He would then blackmail his victims to self-harm or abuse others, threatening to send the compromising images to their friends and family if they did not comply. He traded the abuse material on ‘hurt core’ forums on the Dark Web dedicated to the discussion, filming and image sharing of rape, murder, sadism, paedophilia and degradation.
(2) A case summary provided by Sophie Mortimer of South West Grid for Learning 573to our separate project looking into Intimate Image Abuse:
The Revenge Porn Helpline have been supporting the National Crime Agency on a significant case for over 18 months. One main offender groomed, bribed and blackmailed both children and vulnerable young women to share intimate content. After they had shared relatively mild images, they were then threatened and coerced into sharing ever more extreme images. What started with topless images, moved through masturbation; self-harm; degrading words written on bodies; hitting and hurting themselves, urinating, defecating and even eating their own faeces. The content was all recorded to be sold and shared further online. The NCA believe there are approximately 200 victims and the Helpline is supporting nearly half of these. The Helpline have so far removed 68,441 individual images, currently a 92% takedown rate.
The offender has pleaded guilty to over 100 charges relating to the grooming, coercion and blackmail of the victims.
7.13 As set out in detail in Part 2 below, while consultees broadly echoed our concerns with the term “glorification”, there was strong support for a narrow offence that addressed malevolent encouragement of self-harm. Before we analyse their responses, we set out the current legal position and issues that arise.
7.14 The offences contained in the Suicide Act 1961, Serious Crime Act 2007 (“SCA 2007”) and Offences Against the Person Act 1861 (“OAPA 1861”) are all relevant to considering criminalising encouragement of self-harm. They were considered in both our 2018 scoping report 574and the consultation paper.
7.15 Throughout this chapter we use the term “inchoate offence”. This refers to situations where a substantive offence may not be complete, but due to the conduct that has occurred, the law recognises that a different type of offence has occurred. Put another way, they are standalone offences of encouraging or assisting other conduct. These types of offences were the subject of our 2006 report on Inchoate Liability for Assisting and Encouraging Crime. 575Our recommendations in that report are reflected in Part 2 of the SCA 2007. Inchoate offences of encouragement relate to the interactions between the people involved in criminal activity. For example, where D encourages P to murder V, but V is arrested by police on another matter just as P is about to leave home to murder V. No offence of murder has been attempted or committed. However, D is guilty of an inchoate offence of encouragement of murder.576
7.16 The SCA 2007 creates three inchoate offences that relate to encouraging or assisting crime.577 As observed in Blackstone, it does not matter “whether any ‘anticipated offence’ is ever committed by the other person(s), nor does it matter whether anyone was in fact assisted or encouraged by D’s act.” 578Relevantly, the question of whether a perpetrator’s act was capable of “encouraging or assisting” is a question of fact for determination by the tribunal of fact (the jury or magistrates). 579Further, if the offence being “encouraged or assisted” is itself an inchoate offence, section 49 of the SCA 2007 requires specific intention of encouraging or assisting its commission.580
7.17 In the consultation paper we considered an untested argument that the intentional infliction of self-harm is currently criminalised by the OAPA 1861 The offence of intentionally causing grievous bodily harm, or wounding, contrary to section 18 OAPA
1861 could apply to self-harm. The offence criminalises harm (of the relevant severity) “to any person”, in contrast to section 20 of the same Act that only criminalises harm to “any other person”.581
7.18 If, in line with the above argument, it is inferred that section 18 OAPA 1861 is intended to apply to the causing of really serious bodily harm to oneself, then the provisions of the SCA 2007 may also apply to anyone seeking to assist or encourage such behaviour. This, as we set out in the consultation paper, could apply in the absence of harm being caused to oneself to assistance or encouragement online or offline.582 However, this relies on arguing that self-harm is a crime without any express legislative provision to that effect. This is troubling, particularly in the context of the decriminalisation of suicide in the Suicide Act 1961.583
7.19 The argument based on the OAPA 1861 is potentially problematic for a number of reasons,584 in particular, its reliance on an as yet untested ambiguity, namely the absence of a reference to “any other person” in section 18. This was considered in our scoping consultation paper in that project and referenced in the final report. As we set out in the scoping consultation paper:585
A further difference is that section 18 speaks of wounding or causing grievous bodily harm to “any person”, while section 20 speaks of “any other person”. In this respect section 18 appears to follow the common law offence of mayhem, which forbids the mutilation of either another person or oneself, as in either case this would make the mutilated person unfit to fight for the country.586
7.20 Our concerns with the argument that encouraging self-harm is criminalised by section 18 of the OAPA 1861 and the provisions of the SCA 2007 as set out in the consultation paper remain. In considering addressing the issue of encouragement with a specific offence, it is useful to set out the ways in which the law can capture “encouragement”.
7.21 “Encouragement”, as set out by Smith, Hogan and Ormerod, can encompass a broad range of conduct:587
In Marlow,588 for example, D was convicted on the basis of ‘encouraging’ others by his publication of a book on cannabis cultivation. Encouragement can be by hostile threats or pressure as well as by friendly persuasion. It may be implied as well as express. It was held to be an incitement to advertise an article for sale, stating its potential to be used to do an act which is an offence. It was an act capable of encouraging P to commit that offence589—even when accompanied by a warning that the act is an offence. By contrast, it has been held that merely intending to manufacture and sell, wholesale, a device—which has no function other than one involving the commission of an offence—is not to incite the commission of that offence. 590Arguably, there may be circumstances in which offering such devices for sale would be capable of constituting ‘encouragement’ for the purposes of Part 2 of the 2007 Act.
7.22 The conduct we are trying to capture in the current analysis is related to, but falls below, the threshold of harm in the Suicide Act 1961.
7.23 There is no longer an offence of suicide. “Since there is in law no offence of suicide which can be encouraged or assisted, section 2 [of the Suicide Act 1961] makes the conduct which is capable of encouraging or assisting suicide an offence in its own right.” 591As such, the offence of encouraging or assisting suicide is an entirely inchoate offence, which is committed by doing an act capable of encouraging or assisting suicide whether or not any suicide is attempted or completed. The offence extends to acts capable of encouraging or assisting persons unknown.
7.24 Section 2 sets out the relevant offence:
2— Criminal liability for complicity in another's suicide.
(1) A person (“D”) commits an offence if—
(a) D does an act capable of encouraging or assisting the suicide or attempted suicide of another person, and
(b) D's act was intended to encourage or assist suicide or an attempt at suicide.
(1A) The person referred to in subsection (1)(a) need not be a specific person (or class of persons) known to, or identified by, D.
(1B) D may commit an offence under this section whether or not a suicide, or an attempt at suicide, occurs.
(1C) An offence under this section is triable on indictment and a person convicted of such an offence is liable to imprisonment for a term not exceeding 14 years.
(2) If on the trial of an indictment for murder or manslaughter [of a person it is proved that the deceased person committed suicide, and the accused committed an offence under subsection (1) in relation to that suicide, the jury may find the accused guilty of the offence under subsection (1).
... (4) [...]no proceedings shall be instituted for an offence under this section except by or with the consent of the Director of Public Prosecutions.
7.25 Section 2A extends what can amount to ‘encouraging’ or ‘assisting’ for the purposes of the offence:
2A Acts capable of encouraging or assisting
(1) If D arranges for a person (“D2”) to do an act that is capable of encouraging or assisting the suicide or attempted suicide of another person and D2 does that act, D is also to be treated for the purposes of this Act as having done it.
(2) Where the facts are such that an act is not capable of encouraging or assisting suicide or attempted suicide, for the purposes of this Act it is to be treated as so capable if the act would have been so capable had the facts been as D believed them to be at the time of the act or had subsequent events happened in the manner D believed they would happen (or both).
(3) A reference in this Act to a person (“P”) doing an act that is capable of encouraging the suicide or attempted suicide of another person includes a reference to P doing so by threatening another person or otherwise putting pressure on another person to commit or attempt suicide.
7.26 Section 2B explicitly allows for a course of conduct on behalf of an accused:
2B Course of conduct
A reference in this Act to an act includes a reference to a course of conduct, and a reference to doing an act is to be read accordingly.
7.27 The accused must intend that someone commit or attempt to commit suicide. 592An accused need not know or believe that the person encouraged had been intending or contemplating suicide.593
7.28 The above survey of our analysis in the consultation paper and the law reveals that, while there are specific offences addressing encouraging or assisting suicide, or encouraging or assisting other offences, there is no offence that addresses intentional encouragement of self-harm. In the next section we describe the evidence we received from consultees about the benefits and risks associated with a targeted offence that could address malicious and deliberate encouragement of self-harm.
7.29 In line with the legal and policy positions from the consultation paper summarised in Part 1 above, we asked two specific consultation questions on this topic, one relating to the impact of our provisionally proposed harm-based offence on vulnerable people who post nonsuicide self-harm content, and the other asking whether there should be a specific offence of encouragement of self-harm. We received a mix of general and specific responses that addressed the issue of encouragement of self-harm.
7.30 The consultee responses are summarised and analysed below. In line with the way we posed the consultation questions, the two key issues addressed by consultees were the interaction, if any, between our provisionally proposed harm-based offence and self-harm content, and the possibility of enacting a specific offence of encouragement of self-harm.
7.31 An issue that consultees raised with us in stakeholder meetings as well as in written responses was the phenomenon of online suicide “challenges”. We considered these in the consultation paper, particularly the “Blue Whale Challenge”. South West Grid for Learning (SWGfL), a charity that runs the Professionals Online Safety Helpline, Report Harmful Content and the Revenge Porn Helpline provided a general response jointly with Professor Andy Phippen. They raised concerns with the emphasis we placed in the consultation paper on the “Blue Whale Challenge”, especially the lack of evidence surrounding it and the dangers associated with what they called “digital ghost stories”.594
7.32 The complexities of the “suicide challenges” such as the Blue Whale Challenge were not fully articulated in the consultation paper. Consultees brought this to our attention, and we are particularly grateful to SWGfL and Samaritans for providing extensive responses outlining the challenges that these “digital ghost stories” pose. While these sensitivities exist, many consultees provided responses that outlined the substantial, and problematic, self-harm content that exists quite apart from these types of “challenges”.
7.33 Consultees set out that, although self-harm content online comprises a minority of online content, the most serious examples of it, which may amount to encouragement of self-harm, are not currently criminalised. Responses, including those from Facebook and Samaritans, emphasised that the vast majority of this type of content can be best dealt with through moderation. Facebook described how they do not take down all content discussing suicide and self-harm as it can, amongst other things, “play an important role in supporting people’s recovery from mental health challenges.” They noted the added complexities in dealing with content about eating disorders, for which they have specific policies and remove content as soon as they are made aware of it. More broadly they discussed their mitigation strategies that exist alongside content removal: for example, adding “sensitivity” screen messages, or not allowing content to be discovered via “explore” functions in Instagram.595 This also reflects the concerns of the Government, who noted that vulnerable people seeking help or support should be free to do so online and offline.
7.34 Consultees agreed with our view that the bulk of online content, including much selfharm content, posted by vulnerable people is appropriately treated as a mental health matter.
7.35 Consultation question 28 directly addressed the interaction between our provisionally proposed communications offences and nonsuicide self-harm content:596
Can consultees suggest ways to ensure that vulnerable people who post nonsuicide self-harm content will not be caught by our proposed harm-based offence?
7.36 Selected extracts of consultee responses, and an analysis based on all responses, follows.
7.37 The concerns we raised in the consultation paper about the proliferation of self-harm content online, and the care that must be taken in addressing any encouragement of self-harm, were addressed in a detailed submission we received from Samaritans. They noted:
As we have iterated throughout our response, to avoid the criminalisation of individuals who post self-harm or suicide content, the intent of the user will need to be the primary consideration in prosecutions. Malicious intent without reasonable excuse should be demonstrated in order to meet the threshold for the harm-based intent.
Samaritans also question the Law Commission’s decision to differentiate between non-suicidal self-harm content and suicidal self-harm content - it can be difficult to determine intent to die in regard to self-harm, and it can change over time: self-harm may precede suicide in young people and is considered a strong risk factor in future suicide attempts and suicidal behaviour.
Furthermore, it may be difficult to distinguish between non-suicidal self-harm and a suicide attempt based on online communication alone. To make this distinction may be confusing for those enforcing the offence: an image of self-harm and a suicide attempt using the same method are unlikely to be materially different in appearance and an individual posting content of this nature is likely to be in need of support.
Targeted approaches to individuals may be indicative of malicious intent. For example, a user says they are feeling low on their public social media account, and are then bombarded by unsolicited images of self-harm and information about methods from others, even when asked to stop. The creation of new accounts to bypass blocks to target individuals may indicate malicious intent in some cases, although it should be noted that users may also create multiple new accounts without malicious intent: graphic suicide and self-harm content goes against the community standards of the largest social networks and users who regularly post this content may re-register to avoid detection or evade suspension, neither of which are criminal issues.597
7.38 Refuge noted the importance of protecting survivors of domestic abuse. Their response stressed that any offence, whether a communications offence or a specific offence of encouragement of self-harm, should not inadvertently criminalise women who have posted pictures claiming they related to self-harm when in fact they were victims of domestic abuse:
Some survivors of domestic abuse and other forms of VAWG post descriptions and pictures of the abuse they have suffered as a way of sharing their story and expressing themselves. We have also supported women who have posted pictures of injuries and claim that they were the result of self-harm when in fact they were caused by a domestic abuse perpetrator, in order to discourage others from thinking they are being abused. It is crucial that survivors of domestic abuse are not arrested or prosecuted for communicating about their experience of abuse online.598
7.39 The Association of Police and Crime Commissioners emphasised the importance of CPS guidance and non-criminal regulation. Relatedly, they acknowledged the potential importance of additional policing resources and guidance in referring vulnerable individuals to appropriate resources and support services:
We would suggest that the CPS produces guidance that specifically states that vulnerable people who post non-suicide self-harm content should not be prosecuted under this law. Additionally, College of Policing could produce subsequent guidance encouraging the police not to treat such behaviour as criminal conduct, but when made aware of such behaviour, to refer vulnerable individuals to services that can offer them appropriate support.
As well as the above - or potentially as an alternative - a vulnerable person who posts non suicide self-harm material that is not specifically encouraging others to commit similar behaviour, could arguably be doing so as a form of self-expression, or as an attempt to illicit [sic] emotional support from others to stop the behaviour. Potentially, this could be viewed within the law as proposed as a “reasonable excuse” for the post.
Either way, we believe social media companies have an important role to play in terms of ensuring that users who post self-harm content are signposted to services that can support them.599
7.40 The Magistrates Association observed the tension between providing sufficient protection for those who should not be criminalised and not providing an “automatic defence” for those whose actions do warrant criminal sanction:
As suggested in the consultation, ensuring people who were sharing their own story or struggle, should have a reasonable excuse to the proposed harm-based offence. The challenge is ensuring people can discuss problems (not just relating to themselves, but people may share their experiences relating to supporting friends of family) without providing an automatic defence for people. It may be that communications relating to the issue of non-suicide self-harm should be more tightly defined to ensure only the most grievous offences are caught by the legislation.600
7.41 Big Brother Watch provided a general response, which expressed concerns about many aspects of the approach in the consultation paper. Regarding self-harm they raised concerns about any offence of “glorification of self-harm”, including that selfharm “strongly corresponds with ill mental-health” and that the problem is not well understood. Their concern relates to the fact that “glorification” “all too often refers to individuals experiencing mental health problems and expressing themselves.. .we do not believe that criminalisation is a reasonable response.”601
7.42 The Criminal Bar Association raised concerns about how any protections might work in practice. They emphasised the difficulty of any approach that tried explicitly to categorise or define a “vulnerable person” and, as many legal stakeholders did, stressed the importance of the rigorous application of the public interest test by prosecutors:602
One possible way we suggest would be to include an explicit exclusionary subsection excluding vulnerable people but, as we also say in answer to Question
29 below, we feel there would be inherent difficulties involved both in defining and determining who fell into the category of a “vulnerable person”.
We are conscious of the fact though that whilst a vulnerable person may share NSSH there is still a risk that a person from the likely audience may view such material and go on to self-harm to suicide, that is to say that even where a vulnerable person posts NSSH material there is no way of controlling how that material or post may be used or viewed by other vulnerable people who may be severely adversely affected by it e.g. to the point of self-harm to suicide. This consideration should not be addressed in the offence itself. The public interest test in any decision to prosecute has proven capable of protecting those individuals in respect of whom there is sufficient evidence to secure a conviction but it would be inappropriate or unfair to prosecute them. The offence itself should not distinguish between who may or may be capable of committing an offence in particular circumstances. A sufficient protection exists in the two part prosecution test.603
7.43 There was broad support amongst consultees for the proposition that vulnerable people who post content that includes self-harm content should be protected as much as possible. Some consultees (for example ARTICLE 19) pointed out that these concerns are a result of the “harm-based” offences’ breadth. Consultees with these concerns argued that this demonstrated the dangers of an approach that contains a subjective standard of harm.
7.44 However, many consultees who supported the provisional proposals provided responses that attempted to focus on ways better to protect potentially vulnerable people. The two key elements that recurred across stakeholder responses were the importance of the mental element in any offence, and having robust prosecutorial and police guidance on the matter. As noted in Chapter 6: in relation to cyberflashing, our preference will remain that, as a way of restricting the scope of the offence, guidance and discretion are a “last resort”; in general, we should avoid criminalising conduct that does not warrant criminalisation rather than relying on guidance and discretion to exempt such conduct.
7.45 As we made clear in the consultation paper, the issue of self-harm is quite clearly one that should have strong public health and mental health resourcing. Any response by the criminal law should be targeted and accompanied with safeguards to ensure that the impact on vulnerable people is not disproportionate. Neither our provisionally proposed “harm-based” offence nor a specific encouragement offence should criminalise individuals who are expressing their own experiences with self-harm without any intention to encourage others to self-harm.
7.46 Samaritans rightly raise the lack of a bright-line distinction between suicide and nonsuicide self-harm. In Part 1 above, we have considered the provisions of the Suicide Act 1961 in greater detail than we did in the consultation paper. We acknowledge that there is not a clear distinction in the types of content and behaviours that relate to suicide and self-harm. However, as consultees set out for us in response to Consultation question 29 (analysed below), there appears to be a gap between what would appropriately be addressed by a communications offence and conduct that fails to meet the threshold in the Suicide Act 1961. We aim to address that gap and the most serious examples of encouragement of self-harm.
7.47 Regarding the fault element, many stakeholders emphasised the importance of either requiring specific intention to cause harm or including any potential vulnerability within the “reasonable excuse” element of the provisionally proposed harm-based offence. While we appreciate that it may be a fine distinction, it is important to delineate between an excuse based on individual circumstances and a justification such as reasonable excuse. 604While they may generate sympathy, with good reason, it is conceivable that many individual circumstances may not constitute a “reasonable excuse” in the way that some consultees envisage. In our view a robust fault element provides the best protection to ensure that vulnerable people producing content that is primarily directed to their own experience or seeking help from others are not criminalised.
7.48 Stakeholders (including the Crown Prosecution Service, Association of Police and Crime Commissioners and National Police Chiefs’ Council) also raised the importance of prosecutorial and police guidance alongside any safeguards built in to the offence itself. It was pointed out that whatever safeguards exist in the elements of an offence, having a final layer of discretion to ensure that vulnerable people are not inappropriately prosecuted was vital. Despite this it is worth acknowledging that in some circumstances vulnerable people may well still act in a culpable manner.
7.49 In discussing the issue, the Crown Prosecution Service referred to the matter of R v Gordon.605 In that case a vulnerable defendant was convicted and sentenced under the Suicide Act 1961. The applicant was sentenced to four years’ imprisonment. The Court of Appeal refused leave to appeal against the sentence. While the applicant was acknowledged to have mental health issues and vulnerabilities, the Court held that these were taken into account and given appropriate weight by the sentencing judge. The matter provides an example of culpable conduct engaged in by a vulnerable person, and how an offence that provides sufficient flexibility for prosecutors and sentencing judges can ensure that these vulnerabilities are properly taken into account. In our view, it is important to acknowledge that vulnerable people are still capable of culpably harming other vulnerable people, and to design laws that can be applied appropriately in each circumstance.
7.50 The other related issue discussed by stakeholders is the role of platforms in moderating this type of content. While the criminal offences provide the basis for liability, no doubt the Online Harms regulatory regime will have to deal with content that may not be criminal but remains harmful. 606As noted by the Government in their response to the consultation paper, the Online Harms framework will exist alongside the criminal law. Criminal sanction is just one of the tools (and likely one of “last resort”) that can be used to address this extremely serious problem.
7.51 The other central issue covered by consultees related to a specific offence of encouragement of self-harm. After setting out the issues in the consultation paper, which we summarised above, we sought consultees’ views on a potential offence with built-in safeguards. Consultation question 29 asked:607
Should there be a specific offence of encouragement of self-harm, with a sufficiently robust mental element to exclude content shared by vulnerable people for the purposes of self-expression or seeking support? Can consultees provide evidence to support the creation of such an offence?
7.52 Selected extracts of the consultation responses, and an analysis based on all responses, follows.
7.53 Samaritans agreed. They emphasised the need to avoid stigmatising those who are simply seeking support, and not to hamper the ability of peer support groups, charities and helplines who provide harm minimisation guidance. They also commented on potential benefits of ensuring accountability for those who encourage self-harm with malicious intent. We set out their detailed comments:
As the Law Commission has identified, there are complexities in ensuring people engaging in self-harm are not caught within the proposed offence of encouragement of self-harm. Recent research by Samaritans in England with people who had selfharmed suggested that current provision and support after self-harm was largely unhelpful, and participants reported facing ongoing stigma, meaning that people who have self-harmed may go to informal online spaces to seek peer support without fear of judgement from people they know face to face. Some people who have selfharmed consider professional services to be too basic and want what they consider to be more responsive help in user-led and run online spaces. Capturing users of self-harm communities within this offence would service to further stigmatise individuals who self-harm and may remove an invaluable source of support, without providing an alternative.
It is important to note that peer support may not always be helpful: communities with a focus on self-harm may offer advice on methods of self-harm, guidance on concealing evidence and injuries, and may inculcate the belief that self-harm is an effective coping strategy, which could deter seeking support from other avenues. Furthermore, self-harm peer support groups, helplines and charities may provide guidance on harm minimisation for individuals who are intent on self-harming. An example of this is Self-Injury Support’s resource on Harm Minimisation, which provides guidance on the lowest risk areas to self-harm and the importance of clean implements. This resource is aimed at individuals who are intent on self-harming regardless of intervention and should not be within the scope of the offence: the inadvertent criminalisation of harm reduction approaches from professional organisations would deter the provision of such guidance due to fear of reputational damage and prosecution, potentially leading vulnerable users towards harmful material online. It will be necessary to have the ability to identify bad actors breaking policy within supportive services while also ensuring that service policies designed to keep people safe aren’t inadvertently caught by the law.
Some social media platforms attract a high number of users who share graphic, glamorising images of their self-harm: a recent example seen by Samaritans was of a user sharing an image of a bucket of blood and images of recent self-harm. This attracted responses from other users, remarking on how ‘cool’ this was. Under the same hashtag, users comment on the neatness of others’ injuries, asking about the method used. Many of these users are likely to be young people: self-harm posts on social media often include reference to being caught by their parents, or having to hide paraphernalia relating to self-harm from their parents. This behaviour shows the complexity around assisting self-harm: the content is likely to be harmful to others and shows that users are experiencing feelings of distress, the communication is taking place between minors who are likely to have shared experiences of self-harm and intention may be difficult to understand.
In recent years, multiple online suicide and self-harm challenges, such as Blue Whale, have risen to prominence, and indicate that the overt encouragement of selfharm by bad actors is becoming more of an issue. Media reports and discussions on social media about these challenges can inadvertently raise their profile, encouraging individuals to actively seek them out and potentially cause themselves harm. In the case of Jonathan Galindo, a challenge that emerged in the summer of 2020 and was widely described in the media as the ‘new Blue Whale’, multiple accounts on social media purporting to be this character were created, suggesting that the high level of coverage led to copycat behaviour. Many of these accounts broke the community guidelines of the platforms on which they were created, sharing graphic images of self-harm and injury, and sending private messages to users with the intent of causing them harm.
Another ‘suicide challenge’, the Momo Challenge, received worldwide media attention in 2018. There is no evidence to suggest that the Momo Challenge resulted in actual harm and it is widely considered to be a hoax. However, it spread rapidly via social media, including by high profile celebrities trying to ‘raise awareness’ of the dangers it presented. Samaritans advise that specific details about suicide challenges should be limited to avoid inadvertently drawing attention to potentially harmful content. However, individuals who aren’t vulnerable who create content encouraging self-harm and suicide with malicious intent, whether a hoax or not, should be held accountable for these actions.
Our understanding is that there is no precedent in the UK for prosecuting individuals involved in the creation and spread of online suicide challenges, and we support the prosecution of bad actors who maliciously and deliberately encourage self-harm and suicide. The scale of encouragement to self-harm online is not widely known, perhaps due to the lack of legislation in this area, and while this offence would not be applicable to most individuals speaking about self-harm online, the existence of suicide challenges indicates that there are cases where it would be appropriate. As there is currently little recourse for those encouraging self-harm, the offence may go some way to deter users from creating online challenges in future.608
7.54 The Government agreed that a specific offence would be appropriate. They echoed concerns that were raised in the consultation paper about the potential for any offence inadvertently to criminalise vulnerable people who were sharing content “for the purposes of self-expression or seeking support.” They expressed an interest in ensuring that any offence does not unfairly capture “vulnerable users who may share such content to seek help and support.”609
7.55 The Criminal Bar Association disagreed and emphasised the difficulties involved in such a delicate area of law. However, while their concerns largely mirror our own in the consultation paper, they did not specifically address how safeguards including a higher threshold of harm or a robust fault element could assist in ensuring any offence is appropriately directed:
No. In our view, there should not be a specific offence of encouragement of selfharm. We feel there would be inherent difficulties involved in both defining and determining who fell into the category of a vulnerable person sharing or who had shared content for the purpose of self-expression or seeking support. We also are of the view that these difficulties may lead to inconsistent charging decisions and or outcomes of trials.
This is a complex area and it is unlikely to be dealt with consistently, fairly and properly by reference to the terms of an offence. Each case will turn on its own facts and where a line is crossed or not will be one of judgment and so we advise against over complicating the offence (or any specific additional offence) with prescriptive drafting, which would most likely be incapable of universally fair application.
The proposed offence specifically requires regard to be had of the context. This, in our view, along with the discretionary prosecution test, availability of inchoate offences and any accompanying guidance offers sufficient protection.610
7.56 The Association of Police and Crime Commissioners responded “Other”. They discussed the appropriately limited role of policing in prevention of self-harm behaviour:
The APCC would need to see evidence to support the creation of such an offence before forming a view. We can see arguments that creating a specific offence of encouragement of self-harm may help to deter people from specifically posting content of this nature (as opposed to content which expresses their own experiences with self-harm, or is an attempt to illicit [sic] emotional support from others to stop the behaviour).
However, we believe that conduct such as this will likely come under the proposed harm-based offence as detailed above.
Additionally, we share the Commission’s concerns detailed at paragraph 6.197 in the consultation document, that there is a significant overlap between “victims” and “perpetrators” in this respect: many of those who communicate online about selfharm tend to post content about their own practice and experiences.
Ultimately, policing will not prevent people from engaging in self-harming behaviours. Primarily, a public health approach should be taken to prevent people from harming themselves, or from encouraging others to do the same.611
7.57 A number of consultees responded “Other” and expressed their concerns, largely mirroring those we raised in the consultation paper, about the potential for any offence to criminalise people who may be suffering from mental health problems. These included the responses of:
(1) the Law Society of England and Wales, who commented on the potentially “fine line” between a message that may be a clumsy or misguided attempt to support another and a malicious act of encouragement:
On the one hand we accept that the encouragement of self-harm over the internet is a significant social problem that the law needs to address, and there may well be a case for criminalising it. However, as we raised in a meeting with Commission officials, we are concerned that doing so may criminalise people suffering themselves with mental ill-health. Those considering self-harm and those encouraging them are all likely to be suffering from the same mental health problems. It appears there is a fine line here between people who might inappropriately support each other with such ideas and the malicious person who ‘eggs’ a mentally disordered person to self-harm. Great care must be exercised before criminalising this conduct.612
(2) the Bar Council emphasised that prosecutorial guidance will play an important role should an offence be enacted:
We think it is difficult to create a specific offence of encouragement of selfharm in such a way as to exclude content shared by vulnerable people for the purposes of self-expression or seeking support. In practice, were such a specific offence to be introduced, the Crown Prosecution Service would have to apply its guidance for Crown prosecutors when considering whether the charging standard is met. The Bar Council is not in a position to provide evidence.613
(3) ARTICLE 19 raised the importance of not deterring those who need help from being able to seek it out:
ARTICLE 19 would urge extreme caution in this area. To begin with, it is unclear that self-harm is an offence in and of itself. As the Law Commission rightly notes, criminalisation in this area could result in preventing vulnerable people from seeking help or sharing experiences with others who suffer from the same issues. It is highly unclear that it would help from a medical or mental health perspective.614
(4) the Magistrates Association, along similar lines to the Law Society, noted the potential dangers of any offence that does not have sufficient safeguards to avoid capturing those who may be sending messages of support:
... there are strong reasons against any new offence around encouragement of self-harm. It may be difficult to clearly delineate between supportive messages, and a communication that encourages self-harm with no good intent. It would also be difficult to ensure vulnerable people, especially those sharing their own experiences, would not be captured by any such offence.615
7.58 A range of the comments that accompanied the responses confirm that the caution we expressed in the consultation paper, and explored in Part 1 above, is warranted. However, the majority of consultees supported the creation of an offence of encouragement of self-harm.
7.59 Many consultees stressed that any offence would need to be strictly bounded, both in its drafting and in its application. To that end, consultees discussed the importance of only targeting “bad actors” and ensuring that police and prosecutors are aware of the potential of any offence to impact particularly vulnerable people.
7.60 The potential impact on vulnerable people was used by some consultees as part of arguments made against the project on a broader level. For example, Fair Cop, Legal Feminist and the Free Speech Union argued that the potential impact on vulnerable people illustrated the pitfalls of regulating speech or communication, particularly online communication. However, these responses did not address the very similar justifications that exist for the offence in the Suicide Act 1961 or the arguments that other stakeholders put forward for ensuring that “bad actors” who incite or encourage self-harm are held to account.
7.61 The Criminal Bar Association and Association of Crime and Police Commissioners both registered their concerns with any prospective offence. Each noted the ability of the “harm-based” communications offence 616to target much of the behaviour in a way that takes into account the context of communications, without delving into the particularly delicate issue of self-harm.
7.62 We agree that any encouragement offence must have safeguards built in to ensure similar protection for vulnerable people. However, the matter of Gordon617 referred to by the CPS (albeit in relation to encouraging or assisting suicide) demonstrates how some perpetrators will engage in culpable conduct, which is appropriately addressed by the criminal law, despite their own vulnerability and potential harm to themselves.
7.63 Alongside their response to this consultation question extracted above, Samaritans, confidentially, provided a “rapid review” of relevant online content and accompanying note. These examined evidence of self-harm and suicidal content online and outlined Samaritans’ experience in working with vulnerable people in this area.
(1) These provided nuanced analysis of the various issues, including the positive outcomes for some people in sharing their own experiences of self-harm, and the resources that charities and support organisations provide to attempt to minimise harm to those acutely at-risk.
(2) They note in their consultation response extracted above that while it appears that some “bad actors” do exist, their communications constitute only a small minority of this type of content. Further, that so-called “suicide challenges” like the “Blue Whale” challenge, “Momo” challenge and “Jonathan Galindo” challenge often operate as “myths”618 . Any report or commentary that highlights these events often only draws more attention to them and can in fact lead to a surge of interest in and searches for them.619
7.64 Much of the concern expressed by consultees comes down to the difficulty in delineating between messages of support exchanged between vulnerable people and messages that “encourage” in a culpable way. This is an obvious and real concern. However, these consultees by and large do not consider the possibility of an offence that has an appropriately narrow scope, for example, an offence that only targets intentional “encouragement” or “incitement” to a level of harm approximately equivalent to grievous bodily harm, or suicide. To that end, many of the concerns about vulnerable people would be more convincing if the provisional proposal were broader. For instance, if it was directed at any level of “self-harm” or targeted “glorification” rather than “encouragement” (though the scope of any “encouragement” will remain a live issue).
7.65 Many stakeholders expressed concern at self-harm content in a far broader sense than contemplated in the consultation paper. It will be important to acknowledge that there is a substantial role for any regulator to play, as well as acknowledging the work that many online platforms also do. Any criminal offence will only apply to a very narrow subset of “bad actors” deliberately encouraging or assisting serious self-harm.
7.66 The concerns expressed by consultees may, as considered above, be addressed in large part by ensuring that any offence has a narrow scope. CPS policy and police
education could also accompany any legislative reform. Further, as raised by the NPCC in the context of encouraging or assisting suicide, it is also worth considering whether prosecutions should require the consent of the Director of Public Prosecutions (“DPP”). Similarly, the DPP may consider that the sensitivities of this area could warrant the formulation of a Policy for Prosecutors similar to that issued in respect of cases of encouraging or assisting suicide (although this is of course a matter for the DPP). 620While this issue may be problematic when raised in the context of the communications offences (given the volume of offences and potential impact on resources), any offence directed to encouragement of self-harm would be consciously narrow. DPP consent could be a useful way to ensure that any offence not only has a narrow scope of liability but is also accompanied with a heightened level of prosecutorial oversight and discretion.
7.67 As set out in Part 1 above, existing provisions do not provide a coherent or straightforward way of criminalising the encouragement or assistance of self-harm. However, should a specific offence of “encouraging” self-harm be enacted, the potential breadth of the term “encouragement” would still give rise to the concerns we raised in the consultation paper and that consultees elaborated on in Part 2 above. We now consider how best to address these concerns.
7.68 We set out the potential justification for a specific offence in the consultation paper.621
We have particular concerns about a broad offence of “glorification” of self-harm, rather than a narrower offence of encouragement, akin to the offences under the SCA 2007...
Further, there are arguments against the creation of an encouragement offence. First, even if we set aside the possibility of prosecutions that rely on self-harm being a crime under the OAPA 1861, some of this kind of behaviour would be covered by our proposed harm-based offence. Therefore, there may not be a pressing need for a specific offence, in addition to the proposed harm-based offence.
Second, we are conscious that, in relation to this specific kind of communication, there is a significant overlap between “victims” and “perpetrators”. The aforementioned research suggests that many of those who communicate online about self-harm tend to post content about their own practice of and experiences with self-harm. This suggests that it may be more appropriate to treat NSSI content as a public health issue, using strategies other than the criminal law.
That being said, there may be a case for a narrow offence of encouragement or incitement of self-harm, with a sufficiently robust mental element to rule out the kind of NSSI content shared by vulnerable people for the purposes of self-expression or seeking support.
7.69 The consultation responses that we have received have confirmed our view that a narrow offence of encouragement of self-harm would be beneficial. 622The concerns that we raised in the consultation paper about “glorification” were echoed by consultees, who noted the particular sensitivity of this area of law. At a general level, the breadth of glorification (and that it could capture conduct that does not necessarily deliberately or positively urge action on the part of another) cannot be appropriately constrained in a way that addresses our or consultees’ concerns. While encouragement as a matter of law can still encompass a broad range of conduct, there are ways to ensure that breadth is managed and liability is sufficiently narrow. Any criminalisation can be sufficiently targeted to ensure that vulnerable people are not adversely impacted.
7.70 We believe two key constraints should be built in to any offence that addresses encouragement of self-harm:
(1) the defendant’s conduct should amount to encouragement of or assisting selfharm amounting to serious harm, ie grievous bodily harm (whether or not harm actually resulted from the encouragement); and,
(2) the offence should require proof that the defendant intended a specific result, ie self-harm amounting to grievous bodily harm.
Each is expanded upon below.
7.71 As outlined above, the Suicide Act 1961 criminalises those who encourage or assist suicide. Our aim with the recommended offence is to criminalise that encouragement or assistance of self-harm that falls below this harm threshold, but only the most serious examples of that harm. What follows is a consideration of the appropriate harm threshold for any offence of encouraging or assisting self-harm.
7.72 To be clear, we are not proposing to criminalise the self-harm itself. When we talk of the criminal threshold of harm in respect of the new offence, what we mean is the level of harm that needs to have been encouraged or assisted.
7.73 The criminal law addresses harm in various ways, often responding to different levels of harm with specific offences. Perhaps the most obvious and useful examples of this are contained in the OAPA 1861.623 Violent offences against the person are structured in a rough hierarchy reflecting the seriousness of the harm.
(1) At the lowest end of the range of harm are the common law offences of assault and battery. A charge of battery requires no more than unlawful physical touching. For example, a gentle slap may result in a charge of battery.
(2) The next most serious harm is actual bodily harm (ABH), assault occasioning which is an offence under section 47 of the OAPA. Whilst in practice this offence will only be charged for harm that is more serious than can adequately be addressed by a charge of battery, ABH is any hurt or injury that interferes with the health or comfort of the victim and is more than “transient or trifling”.624 As per our Report on Offences Against the Person:
As well as the more obvious and commonplace types of injury such as bruises and grazes, the definition captures a huge range of harms including cutting off a person’s hair and causing a temporary loss of consciousness. It has also been held to include causing a recognised psychiatric condition.625
(3) The most serious category of harm (falling short of death) is wounding or grievous (really serious) bodily harm (GBH). Wounding or causing GBH are offences under sections 18 and 20 of the OAPA (the distinguishing factor between the sections being one of intention: both sections concern wounding and GBH). It should be noted that wounding does not need to amount to GBH in order to be an offence under these sections; wounding refers to a break in all the layers of the skin, whether or not the injury is really serious. This distinction is important in what follows: when we refer to grievous bodily harm, we do not propose to include within that definition wounding not amounting to grievous bodily harm.
7.74 One important feature of these various offences and levels of harm is that consent does not operate equally across each. Whilst lack of consent is an important element in some offences - it is partly what makes the touching unlawful in battery - other offences can be committed regardless of free and willing consent. In relation to more serious harm, such as ABH or GBH, consent is only legally effective if provided within a context accepted in case law as providing a public policy justification for the injury or risk of injury. 626Examples of this would include surgery - a surgeon faces no criminal liability for operating on a consenting patient - or certain sports such as boxing. Outside of the accepted exceptions, however, the fact that a person consented to the harm done to them is of no legal effect.
7.75 The implication of this for our purposes is worth noting. Whilst it would certainly be going too far to argue that all acts of self-harm were acts of free and willing consent to harm - questions of capacity would certainly arise in some contexts - there can equally be little doubt that some acts of self-harm are, by their nature, consensual. If we were simply to criminalise all encouragement of self-harm, regardless of the severity of that harm, this could have the effect of criminalising the encouragement of an act that, when actually committed against another, would be legal. So no battery can be committed by A against willing B - the application of force would be lawful -but A’s encouragement of B’s act of self-battery would be illegal. This would seem contradictory.
7.76 Therefore, the lowest level of harm that our proposed offence could target (quite apart from the question of whether that is the appropriate level) is that harm to which consent is not an invariable determinant of liability, ie ABH.
7.77 In practice, we do not consider that this would pose any real problems regarding the scope of the offence. Our consultees have confirmed our view that the type of harm against which our offence ought to protect is at least serious harm; certainly, it is more than transient or trifling harm.
7.78 Of course, on its own, this fact should not determine the threshold of harm for the offence. Just because harm amounting only to battery is necessarily out of scope, it does not follow that harm amounting to ABH should be within scope. Indeed, we share those concerns of consultees who noted the potential of a broad criminal offence to criminalise victims. We know that many children and young adults, including preteens, discuss their self-harm on social media as a coping mechanism, and it is not clear that this could never amount to encouragement or assistance.627 As we note above, ABH covers harms ranging from bruises, grazes and scratches to more serious harms such as temporary loss of consciousness. This would cover a very broad range of self-harm - and in so doing bring within scope a great many more young and vulnerable people - including some of those harms that are recommended as safer alternatives to more serious forms of self-harm.
7.79 This is not to say that we should be unconcerned about encouragement of self-harm amounting to ABH. Rather than asking about a false dichotomy between harms that are supposedly “criminal” from those that inherently “not criminal” - as we see in offences against the person, all harms can be criminal - the better question in protecting people from self-harm is at what point public health campaigns should cede ground to the criminal law.
7.80 It is for a similar reason that we propose to focus on grievous bodily harm rather than wounding as the right threshold to establish the offence. As we note above, wounding is treated in the OAPA 1861 as equivalent to GBH in terms of seriousness, whether or not the wounding amounted to GBH. We do not consider this to be an appropriate equivalence in the realm of self-harm. Alongside poisoning, wounding is one of the paradigmatic forms of self-harm, yet some wounding (such as cutting) does not amount to grievous bodily harm. Therefore, to include it in the offence would necessarily bring within scope encouragement or assistance of self-harm that is not of sufficient seriousness to justify a criminal law response.
7.81 Finally, we note that disordered eating formed part of the potentially harmful NSSI content we considered in the consultation paper and that consultees addressed in their responses. It may need to be a particularly extreme example of disordered eating to meet the threshold of GBH. By way of illustration, a perpetrator could encourage another to provoke vomiting with an implement. However, such examples would be captured by this proposed threshold. Further, content relating to knowingly false dietary advice that causes harm may well be captured by the separate offence we propose to deal with the deliberate harmful communication of false information.
7.82 Our aim is only to criminalise the most serious examples of encouragement of selfharm. This has been partly addressed above by setting a very high threshold of harm. However, the other important constraint is ensuring that only those offenders with a particular intention are captured.
7.83 In our view the fault element of an offence of encouragement of self-harm should require that the offender intend that their actions encourage another to self-harm amounting to grievous bodily harm. This reflects the position we initially set out in the consultation paper, and had confirmed by consultees, that there are instances of intentional encouragement of self-harm to the level of grievous bodily harm. It also ensures that only those who intend to encourage or assist serious self-harm will be culpable. By way of counterexample, where a defendant intends another to harm themselves amounting only to actual bodily harm, but grievous bodily harm results, the defendant would not have the requisite fault.
7.84 We believe this fault element is appropriate in the circumstances for two reasons:
(1) it mirrors the actus reus or conduct element of the offence, being encouraging or assisting self-harm amounting to at least grievous bodily harm; and,
(2) requiring this level of intention will avoid making criminals of either those who discuss or depict their own self-harm (possibly as a coping mechanism or warning) but with no intention that others follow suit, or those who provide advice about methods of self-harm that avoid serious injury.
7.85 The second reason above requires some qualification. While we are confident that the elements we recommend will generally constrain the offence to only the most culpable offending, there may be some rare instances where a person A assists an act of GBH knowing that the self-harm will occur, albeit that their purpose was to avoid an even worse outcome (such as even more serious injury or death). Such actions would satisfy the elements of the offence.
7.86 For example, a person A may provide clean razor blades to another person B in the full knowledge that B will injure themselves in any event, albeit that, absent clean blades, B may use dirty implements or other yet more dangerous mechanisms to harm themselves. A is assisting B’s inevitable act of self-harm, though their purpose is to mitigate that harm.
7.87 Examples such as these are extreme: certainly, the provision of razor blades to another person who had never expressed any inclination to self-harm, or to a vulnerable person with no attempt to help them avoid self-harm or otherwise mitigate their conduct, may well be appropriately culpable conduct. What renders this example relatively extreme is the inevitability of the self-harm; serious injury or even death were unavoidable.
7.88 In certain cases like those described immediately above, prosecutions may still be considered in the public interest. While every matter will turn on its own facts, we note that positive defences may have a role to play. For example, the defence of necessity (duress of circumstances), operates to ensure that actions undertaken in such extraordinary circumstances proportionately to avoid consequences or mitigate harm should not be criminal. 628It is a feature of the criminal law that certain acts that would otherwise be crimes can be justified on the basis of some other factor: for example, the intentional infliction of even serious injury may be justified on the basis of selfdefence. So it is with necessity: in some situations, a person may be constrained by the circumstances to act as they did in order to avoid, for example, death or serious injury (or, in this case, more serious injury).
7.89 Further, while our main concern is ensuring that the scope of any offence is appropriately narrow, edge cases such as these will clearly raise complex public interest matters that would have to be considered by the CPS when determining whether prosecution is appropriate. We consider this further in the following section.
7.90 Another matter we have considered is the maximum penalty for any new offence of encouraging or assisting self-harm. In our view, given the very serious threshold of harm, and the strict fault element, the maximum penalty should reflect the greater culpability of those convicted of this offence than the general harm-based communications offence we recommend in Chapter 2: . However, we also accept that the conduct captured by this offence is comparatively less serious than that captured by the offence in section 2 of the Suicide Act 1961, which is an indictable offence with a maximum penalty of up to 14 years’ imprisonment. In our view, a new offence of encouraging or assisting self-harm should be an indictable offence with a penalty that reflects the seriousness of the culpable conduct proportionate to the more serious conduct that would be captured under section 2 of the Suicide Act 1961.
7.91 The considerations above all relate to the criteria for liability: the conduct and fault elements of any offence of encouragement of self-harm. However, as many consultees stressed, any offence must also be accompanied with adequate safeguards in enforcement. As we noted in the consultation paper, 629and as many consultees emphasised,630 the vast majority of NSSI content may be more appropriately addressed as a public health issue.
7.92 Many consultees, including the Crown Prosecution Service itself, noted the importance of CPS policy and prosecutorial discretion in ensuring that acts of vulnerable people are not inadvertently criminalised. 631We agree, and believe that the DPP should consider issuing a specific CPS policy for any offence of encouragement or assistance of self-harm similar to the existing policy for encouraging or assisting suicide.632
7.93 The edge cases discussed at paragraph 7.87 above will likely demand careful consideration of the public interest. We acknowledge that, in certain circumstances, it is appropriate that conduct should not be prosecuted albeit that it might fall within the parameters of an offence, and note the expertise that specialist prosecutors have in taking public interest decisions in these circumstances. The CPS policy for encouraging or assisting suicide in relation to the public interest sets out factors which make prosecution more or less likely to be in the public interest. The following such factors could also be useful in the context of an offence of encouraging or assisting self-harm:633
A prosecution is less likely to be required if:
1. the victim had reached a voluntary, clear, settled and informed decision to commit suicide;
2. the suspect was wholly motivated by compassion;
3. the actions of the suspect, although sufficient to come within the definition of the offence, were of only minor encouragement or assistance;
4. the suspect had sought to dissuade the victim from taking the course of action which resulted in his or her suicide;
5. the actions of the suspect may be characterised as reluctant encouragement or assistance in the face of a determined wish on the part of the victim to commit suicide; ...
7.94 We have considered whether a different fault element could obviate the requirement for such a policy. However, an offence with a narrower fault element would fail to capture considerable culpable conduct and we are not convinced that it would be practically workable. Even if we were to import other elements into the offence, such as proportionality, these would inevitably closely mirror the elements of the existing defences (such as duress of circumstances). Not only would this add little to the law, it would introduce real complexity into every prosecution, even when it is clear that such exempting factors might be irrelevant. Further, the approach that we recommend mirrors that taken in the Suicide Act 1961, and to adopt a different fault element would not encourage consistency of approach between these provisions.
7.95 It is of primary importance that any offence does not over-criminalise. However, we recognise that no matter the safeguards “built-in” to the elements of an offence, some conduct may be unintentionally caught. As emphasised in Part 2 above, police and prosecutorial decisions will be key to ensuring that any such conduct is properly dealt with.
7.96 Alongside the high threshold for harm and the deliberately narrow scope of the offence, we think it is appropriate to include a requirement that consent of the Director of Public Prosecutions be required,634 and that the Director of Public Prosecutions consider introducing a policy similar to the one that exists concerning prosecutions under the Suicide Act 1961.
7.97 We recommend that encouraging or assisting serious self-harm should be an indictable offence with the following elements:
(1) A person (A) intentionally encouraged or assisted another person (B) to cause himself or herself serious physical harm.
(2) B need not be a specific person (or class of persons) known to, or identified by, A.
(3) A may commit this offence whether or not serious physical harm, or an attempt at causing serious physical harm, occurs.
(4) Physical harm is serious if it would amount to grievous bodily harm under the Offences against the Person Act 1861.
7.98 We recommend that instituting proceedings for the offence of encouraging or assisting serious self-harm should require the consent (but not personal consent) of the Director of Public Prosecutions.
7.99 We recommend that the Director of Public Prosecutions consider issuing a policy for the offence recommended in Recommendation 14 above similar to the one that exists concerning prosecutions under the Suicide Act 1961.
8.1 We recommend that section 127(1) of the Communications Act 2003 and the Malicious Communications Act 1988 be replaced with a new offence with the following elements:
(1) the defendant sent or posted a communication that was likely to cause harm to a likely audience;
(2) in sending or posting the communication, the defendant intended to cause harm to a likely audience; and
(3) the defendant sent or posted the communication without reasonable excuse.
(4) For the purposes of this offence:
(a) a communication is a letter, article, or electronic communication;
(b) a likely audience is someone who, at the point at which the communication was sent or posted by the defendant, was likely to see, hear, or otherwise encounter it; and
(c) harm is psychological harm, amounting to at least serious distress.
(5) When deciding whether the communication was likely to cause harm to a likely audience, the court must have regard to the context in which the communication was sent or posted, including the characteristics of a likely audience.
(6) When deciding whether the defendant had a reasonable excuse for sending or posting the communication, the court must have regard to whether the communication was, or was meant as, a contribution to a matter of public interest.
8.2 We recommend that “likely” be defined as “a real or substantial risk”.
8.3 We recommend that the offence be an either-way offence, with a maximum
sentence comparable to the existing Malicious Communications Act 1988.
Paragraph 2.257
8.4 We recommend that the harm-based communications offence applies to communications that are sent or posted in England and Wales, or that are sent or posted by a person habitually resident in England and Wales.
Paragraph 2.274
8.5 We recommend that the existing offences in sections 127(2)(a) and (b) of the Communications Act 2003 be replaced with a new summary offence targeting knowingly false harmful communications with the following elements:
(1) the defendant sent or posted a communication that he or she knew to be false;
(2) in sending the communication, the defendant intended to cause non-trivial psychological or physical harm to a likely audience; and
(3) the defendant sent the communication without reasonable excuse.
(4) For the purpose of this offence:
(a) a communication is a letter, electronic communication or article (of any description); and
(b) a likely audience is someone who, at the point at which the communication was sent by the defendant, was likely to see, hear, or otherwise encounter it.
8.6 We recommend the offence should have a limitation period of three years from the day the offence was committed and six months from the day the prosecution becomes aware of sufficient information to justify proceedings.
Paragraph 3.71
8.7 We recommend that the offence in section 127(2)(c) of the Communications Act 2003 be replaced with a summary offence to address hoax calls to the emergency services.
Paragraph 3.94
8.8 We recommend that a separate either-way offence explicitly addressing genuinely threatening communications be enacted with the following elements:
(1) the defendant sent or posted a communication which conveys a threat of serious harm;
(2) in conveying the threat:
(a) the defendant intended the object of the threat to fear that the threat would be carried out; or
(b) the defendant was reckless as to whether the object of the threat would fear that the threat would be carried out.
(3) The defendant would have a defence if he or she could show that a threat to cause serious financial harm was used to reinforce a reasonable demand and he or she reasonably believed that the use of the threat was a proper means of reinforcing the demand.
(4) For the purpose of this offence:
(a) a communication is a letter, electronic communication or article (of any description); and
(b) serious harm includes serious injury (amounting to grievous bodily harm as understood under the Offences Against the Person Act 1861), rape and serious financial harm.
Paragraph 3.135
8.9 We recommend that the intentional sending of flashing images to a person with epilepsy with the intention to cause that person to have a seizure should be made an offence.
Paragraph 3.142
8.10 We recommend that both the general harm-based communications offence and the knowingly false communications offence have an explicit press exemption. This exemption should not apply to comments made by readers on news articles online.
Paragraph 4.34
8.11 We recommend that the Sexual Offences Act 2003 be amended to include an offence of cyberflashing with the following elements:
(1) The defendant (A) intentionally sent an image or video recording of any person’s genitals to another person (B), and
(2) either-
(a) A intended that B would see the image or video recording and be caused alarm, distress or humiliation, or
(b) A sent the image or video recording for the purpose of obtaining sexual gratification and was reckless as to whether B would be caused alarm, distress or humiliation.
Paragraph 6.133
8.12 We recommend that Schedule 3 of the Sexual Offences Act 2003 be amended to include the offence of cyberflashing so that notification requirements under Part 2 of the Sexual Offences Act 2003 are imposed when an appropriate seriousness threshold is met.
Paragraph 6.134
8.13 We recommend that Sexual Harm Prevention Orders be available for all cyberflashing offences.
Paragraph 6.135
|
8.14 We recommend that victims of the new cyberflashing offence should have automatic lifetime anonymity. |
|
Paragraph 6.136 |
|
8.15 We recommend that victims of the new cyberflashing offence should automatically be eligible for special measures at trial. |
|
Paragraph 6.137 |
8.16 We recommend that restrictions on the cross-examination of victims of sexual offences should extend to victims of the new cyberflashing offence.
Paragraph 6.138
8.17 We recommend that encouraging or assisting serious self-harm should be an indictable offence with the following elements:
(1) A person (A) intentionally encouraged or assisted another person (B) to cause himself or herself serious physical harm.
(2) B need not be a specific person (or class of persons) known to, or identified by, A.
(3) A may commit this offence whether or not serious physical harm, or an attempt at causing serious physical harm, occurs.
(4) Physical harm is serious if it would amount to grievous bodily harm under the Offences against the Person Act 1861.
Paragraph 7.97
8.18 We recommend that instituting proceedings for the offence of encouraging or assisting serious self-harm should require the consent (but not personal consent) of the Director of Public Prosecutions.
Paragraph 7.98
8.19 We recommend that the Director of Public Prosecutions consider issuing a policy for the offence recommended in Recommendation 14 above similar to the one that exists concerning prosecutions under the Suicide Act 1961.
Paragraph 7.99
Consultees
8.20 We received 132 responses to our consultation. We are extremely grateful for the time that consultees took to share their experiences with us, and the perspectives they offered on our provisional proposals. A large number of responses reflected the substantial time and effort that expert consultees spent analysing and responding to our consultation paper. We are confident that the consultation process strengthened our report.
8.21 The consultation was conducted during the COVID-19 pandemic and associated restrictions. We understand the exceptional pressures that all consultees have been subject to during this time, and are grateful that they still took the time to engage with our consultation.
8.22 Four consultees requested to remain anonymous due to the nature of the online abuse they had suffered.
8.23 We received responses from the following individuals: Barry Thain, Elizabeth Tudor, Wendy Cockcroft, Ken Grayling, Rebecca Carter, Alison Griffiths, P Dore, Sophie Gallagher, Sarah Jones, Adisa Nicholson, Samuel Rowe, Jack Astrom, Prof T Keren-Paz, Rob, Dr R Sewell, Ian Milburn, Flora Courtney, Christian Wilton-King, Christopher Barnes, Araba Taylor, Andy Gregory, Christopher Watson, Tommy Axford, Gina Miller, Sam Guinness, Shereen Thomas, Prof A Gillespie, Dr Z Alhaboby, Jane Ayers, Naa Djaba, Dr J Neller, Michelle Thompson, Dr L Thompson, Rose Venice Allan, Nigel Parker, Laura Caulton, Dr K Barker & Dr O Jurasz, Dr E Tiarks & Dr M Oswald, John Carr, George Clarke, Jonathan West, Nick Podmore, Emily Staite, Ian Millard, Philip Hurst, Ryan McKenzie, Paul Grimaldi, Ronan Fahy, Prof C McGlynn QC (Hon) & Dr K Johnson, Laura Higson-Bliss, Chara Bakalis, Ruth Deech, Prof J Rowbottom.
8.24 We received responses on behalf of the following organisations: The Premier League, Her Majesty’s Government, Facebook, Northern Ireland Department of Justice, Open Rights Group & Preiskel & Co LLP, Full Fact, IMPRESS, Greater Manchester Deputy Mayor, Big Brother Watch, DMG Media, Radical, Board of Deputies of British Jews, Adam Smith Institute, News Media Association, Match Group, Verizon Media, Victims’ Commissioner of England & Wales, Rail Delivery Group, Epilepsy Action, Online Dating Association, SWGfl & Prof A Phippen, Sex Matters, National Police Chiefs’ Council, Criminal Bar Association, Alan Turing Institute, Magistrates Association, Free Speech Union, ARTICLE 19, Criminal Law Solicitors’ Association, Legal Feminist, Bar Council of England & Wales, Suzy Lamplugh Trust, Law Society of England & Wales, English PEN, Hacked Off, Samaritans, Media Lawyers’ Association, Association of Police & Crime Commissioners, Carnegie UK Trust, Changing Faces, National Secular Society, Antisemitism Policy Trust, Angelou Centre, Mermaids, Women’s Aid, Epilepsy Society, Crown Prosecution Service, #NotYourPorn, End Violence Against Women Coalition, Merched Cymru, Stonewall, NPPC LGBT+ Portfolio, Fix the Glitch, National AIDS Trust, Refuge, Demos, Friends Families & Travellers, Zoom, Wild Woman Writing Club, British Transport Police, LGB Alliance, Young Epilepsy, Judicial Security Committee, Fair Cop, Keep Prisons Single Sex, Kingsley Napley, Justices’
Legal Advisers’ and Court Officers’ Service, Derbyshire Constabulary, Townswomen’s Guilds, Community Security Trust, Leonard Cheshire, LGBT Fed, Proverbs 31.
Glossary
4chan
4chan is a website to which images and discussion can be posted anonymously by internet users. The website contains a number of sub-categories - or “boards” - such as, notably, the “Politically Incorrect” board and the “Random” board. The website has proved controversial, and has at times been temporarily banned by various internet service providers.
App
Short for “application”, this is software that can be installed on a mobile device, such as a tablet or mobile phone, or a desktop computer.
AirDrop
This is an Apple service that allows users to transfer files (including photographs) between Apple devices using a peer-to-peer wireless connection (ie they are not sent over the internet or mobile network).
Blog
An online journal, or “web log”, usually maintained by an individual or business and with regular entries of content on a specific topic, descriptions of events, or other resources such as graphics or videos. To “blog” something is also a verb, meaning to add content to a blog, and a person responsible for writing blog entries is called a “blogger”. Microblogging refers to blogging where the content is typically restricted in file size; microbloggers share short messages such as sentences, video links or other forms of content. Twitter is an example of a microblog.
Body dysmorphia
This is a chronic mental health condition characterised by extreme anxiety or obsession over perceived physical flaws.
Chatroom
A feature of a website where individuals can come together to communicate with one another.
Chatrooms can often be dedicated to users with an interest in a particular topic. Chatrooms can have restricted access or be open to all.
Comment
A response to another person’s message - such as a blog post, or tweet - often over a social media platform.
Cyberbullying
The use of the internet enabled forms of communication to bully a person, typically by sending messages of an intimidating or threatening nature.
Cyberflashing
The term “cyberflashing” is used to refer to a range of behaviours, but mostly commonly involves a man sending an unsolicited picture of his genitals to a woman.
Cyberstalking
A form of stalking that takes place over the internet.
Dragging
Also known as “Trashing” - the practice of repeatedly publicly posting about a person in order to humiliate or shame them, often on a chatroom or similar social media site.
Defendant
The person accused of an offence.
Dick pic
Strictly speaking, this is photograph that a person has taken of their penis. The term more commonly relates to these photographs being sent to another or posted publicly.
Doxing
Searching for and publishing private or identifying information about a particular individual on the web, typically with malicious intent.
Either-way offence
An offence that can be tried either in the Crown Court or in a magistrates’ court.
A social media platform which connects users from all over the world and enables them to post, share, and engage with a variety of content such as photos and status updates.
Fake news
False, often sensational, information disseminated under the guise of news reporting.
Fault element
A culpable state of mind that needs to be proven beyond reasonable doubt in order to establish criminal liability. Also referred to as “mens rea” or “mental element”.
Friend
The term used on social media services such as Facebook to refer to an individual who is added to a user’s social network on the platform. A person may allow this “friend” to view their profile, or particular parts of it (for example, certain posts or messages). It is also used as a verb, for example, to “friend” a person, means to add them to your social network. Facebook “friends” may not actually be “friends” in the conventional understanding of the term. Someone could “friend” a complete stranger.
Follow
“Following” another user of certain social media platforms (for example, Twitter or Instagram) means that you will receive updates from that user, which will appear in your newsfeed.
GIF
A GIF (“graphics interchange format”) is a moving or “animated” digital image that plays back (or “loops”) continuously. They are mostly soundless, and can include short clips of video or film as well as cartoons.
Hashtag
A hashtag is a tag usually used on social networks such as Twitter or Facebook. Social networks use hashtags to categorise information and make it easily searchable for users. It is presented as a word or phrase preceded by a #. For example, a current well-known hashtag is MeToo.
Hate Crime
There is no statutory definition of “hate crime”. When used as a legal term in England and Wales, “hate crime” refers to two distinct sets of provisions:
Aggravated offences under the Crime and Disorder Act 1998 (“CDA 1998”), which are offences where the defendant demonstrated, or the offence was motivated by racial or religious hostility;
Enhanced sentencing provisions under the Criminal Justice Act 2003 (“CJA 2003”), which apply to offences where the defendant demonstrated, or the offence was motivated by hostility on the grounds of race, religion, sexual orientation, disability or transgender identity.
A different definition is used by the police, Crown Prosecution Service and National Offender Manager Service for the purposes of identifying and flagging hate crime. The focus of this definition is on victim perception:
Any criminal offence which is perceived by the victim or any other person, to be motivated by a hostility or prejudice based on a person’s race or perceived race; religion or perceived religion; sexual orientation or perceived sexual orientation; disability or perceived disability and any crime motivated by a hostility or prejudice against a person who is transgender or perceived to be transgender.
The term hate crime is sometimes also used to describe “hate speech” offences, such as offences of stirring up hatred under the Public Order Act 1986, and the offence of “indecent or racialist chanting” under the Football (Offences) Act 1991.
Inchoate offence
An offence relating to a criminal act which has not, or not yet, been committed. The main inchoate offences are attempting, encouraging or assisting crime.
Indictable offence
An offence triable in the Crown Court (whether or not it can also be tried in a magistrates’ court); contrasted with a summary offence.
A photo sharing app that allows users to take photos, apply filters to their images, and share the photos instantly on the Instagram network and other social networks such as Facebook or Twitter.
Internet Access Provider
A company that provides subscribers with access to the internet.
Internet Service Provider
A broader term than Internet Access Provider referring to anything from a hosting provider to an app creator.
IP address
An “internet protocol” address is a numerical label which identifies each device on the internet, including personal computers, tablets and smartphones.
Liking
Showing approval of a message posted on social media by another user, such as his or her Facebook post, by clicking on a particular icon.
Meme
A thought, idea, joke or concept that has been widely shared online, often humorous in nature - typically an image with text above and below it, but sometimes in video and link form.
Non-binary
An umbrella term for people whose gender identity doesn’t sit comfortably with “man” or “woman”. It can include people who identify with some aspects of binary gender identities, and others who completely reject binary gender identities. Non-binary people may also identify under the transgender umbrella.
Offline communication
Communication that does not use the internet (for example, having a face-to-face conversation or sending a letter).
Online abuse
For the purposes of this report, we adopt the following working definition of “online abuse”. Online abuse includes but is not limited to: online harassment and stalking; harmful one-off communications, including threats; discriminatory or hateful communications, including misogynistic communications (“online hate”); doxing and outing; impersonation.
Online communication
Communication via the internet between individuals and/or computers with other individuals and/or computers.
Online hate
By “online hate” we mean a hostile online communication that targets someone on the basis of an aspect of their identity (including but not limited to protected characteristics). Such communications will not necessarily amount to a hate crime. We note that the College of Policing’s Hate Crime Operational Guidance (2014), stipulates that police should record “hate incidents” using a perception-based approach. Again, such incidents may not amount to a hate crime.
Photoshop
A software application for editing or retouching photographs and images.
Pile-on harassment
Harassment that occurs where many individuals send communications that are harassing in nature to a victim. This can involve many hundreds or thousands of individual messages and is relatively closely connected to the instantaneous and un-geographically bounded nature of the online environment.
Post or posting (on social media)
A comment, image or video that is sent so as to be visible on a user’s social media page or timeline (whether the poster’s own or another’s).
Private message
A private communication between two people on a given platform which is not visible or accessible to others.
Protected characteristics
In the context of hate crime this refers to characteristics that are specified in hate crime laws in England and Wales, namely; race, religion, sexual orientation, disability and transgender status. The term is also sometimes used in the context of the Equality Act 2010, which specifies nine protected characteristics. There is some overlap between the two, but in this report we are referring to the hate crime characteristics unless we specify otherwise.
Replying
An action on, for example, Twitter that allows a user to respond to a Tweet through a separate Tweet that begins with the other user’s @username.
Retweeting
The re-sharing (forwarding) on Twitter by a person (B) of a message received from another person (A), using the re-tweet button and attributing the message to A.
Sharing
The broadcasting by users of social media of web content on a social network to their own social media page, or to the page of a third party.
Skype
A free program that allows for text, audio and video chats between users; it also allows users to place phone calls through their Skype account.
Social media
Websites and apps that enable users to create and share content or to participate in social networking.
Social media platform
Refers to the underlying technology which facilitates the creation of social media websites and applications. From a user’s perspective, it enables blogging and microblogging (such as Twitter), photo and video sharing (such as Instagram and YouTube), and the ability to maintain social networks of friends and contacts. Some platforms enable all of these in one service (through a website and/or an application for a desktop computer or mobile phone) as well as the ability for third-party applications to integrate with the service.
Social Networking Service
A service provided by an internet company which facilitates the building of social networks or social relations with other people, through the sharing of information. Each service may differ and target different uses and users. For example, facilitating connections between business contacts only, or only particular types of content, such as photos.
Summary or summary-only offence
An offence triable only in a magistrates’ court; in contrast to an indictable or either-way offence.
Tag
A social media function used commonly on Facebook, Instagram and Twitter, which places a link in a posted photograph or message to the profile of the person shown in the picture or targeted by the update. The person that is “tagged” will receive an update that this has occurred.
Troll
A person who creates controversy in an online setting (typically on a social networking website, forum, comment section, or chatroom), disrupting conversation as to a piece of content by providing commentary that aims to provoke an adverse reaction.
Tweet
A post on the social networking service Twitter. Tweets can contain plain text messages (not more than 280 characters in the English version of the service), or images, videos, or polls. Users can Tweet to another person (@mention tweets) so as to ensure they will be notified of the Tweet, or can also message them directly. Other users can retweet the Tweets of others amongst their connections on the platform.
A social network that allows users to send “Tweets” to their followers and/or the public at large.
VoIP
“Voice over Internet Protocol” refers to the group of technologies that allow users to speak to each other over an Internet Protocol (such as the internet) rather than over traditional networked telephone services.
An encrypted instant messaging service for one-to-one or group chat on mobile devices.
YouTube
A video-sharing website that allows registered users to upload and share videos, and for any users to watch videos posted by others.
Zoom
Zoom is cloud-based videoconferencing and messaging software that allows for video meetings with multiple participants (up to 1,000 in some cases).
CCS0621857840
978-1-5286-2771-9
Internet Live Stats, https://www.internetlivestats.com/one-second/#traffic-band (last visited 13 July 2021).
The Number of Tweets per Day in 2020, https://www.dsayce.com/social-media/tweets-day/ (last visited 13 July 2021.
WhatsApp Revenue and Usage Statistics (2020), https://www.businessofapps.com/data/whatsapp-statistics/ (last visited 13 July 2021).
Internet Live Stats, https://www.internetlivestats.com/one-second/#traffic-band (last visited 13 July 2021).
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, especially Chapter 4.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, para 4.9.
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, paras 2.155 to 2.161.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No
248, para 4.12.
That is, 61% of respondents selected at least one of the things listed when asked “Which if any of these things have you come across on the internet in the last 12 months?”. The things listed were potential online harms, divided into three categories: (1) data; (2) hacking/security; and (3) content/contact with others.
Under (3), content/contact with others, the following things were listed: (i) fake news; (ii) offensive language; (iii) violent\ disturbing content; (iv) unwelcome friend\follow requests/unwelcome contact or messages from strangers; (v) offensive videos/pictures; (vi) harmful/misleading advertising; (vii) hate speech/inciting violence; (viii) bullying, abusive behaviour or threats; (ix) trolling (a person who deliberately says something controversial); (x) people pretending to be another person; (xi) sexual\pornographic content; (xii) spending too much time online; (xiii) encouraging self-harm e.g. cutting, anorexia, suicide; (xiv) encouraging terrorism\ radicalisation; (xv) cyberstalking (harassment from other internet users); (xvi) material showing child sexual abuse.
Jigsaw Research, Internet users’ experience of harm online: summary of survey research (Ofcom 2019), https://www.ofcom.org.uk/ data/assets/pdf file/0028/149068/online-harms-chart-pack.pdf (last visited 13 July 2021).
H Bentley and others, How safe are our children? 2019: an overview of data on child abuse online (NSPCC 2019), https://learning.nspcc.org.uk/media/1747/how-safe-are-our-children-2019.pdf (last visited 13 July 2021), at p 10.
Glitch and End Violence Against Women Coalition, “The Ripple Effect: COVID 19 and the Epidemic of Online Abuse” (2020) available at: https://www.endviolenceagainstwomen.org.uk/wp-content/uploads/Glitch-and-EVAW-The-Ripple-Effect-Online-abuse-during-COVID-19-Sept-2020.pdf (last visited 13 July 2021).
Alan Turing Institute, “Detecting East Asian prejudice on social media”, available at: https://www.turing.ac.uk/research/research-projects/hate-speech-measures-and-counter-measures/detecting-east-asian-prejudice-social-media (last visited 13 July 2021).
This figure comes from internal data generated by the National Online Hate Crime Hub, provided to us by Paul Giannasi.
See for example, S Stone, “Premier League plans social media blackout in response to online abuse” BBC News 18 April 2021 available at: https://www.bbc.co.uk/sport/football/56792443 (last visited 13 July 2021).
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, para 1.21.
Draft Online Safety Bill 2021, available at:
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment data/file/985033/D raft Online Safety Bill Bookmarked.pdf (last visited 13 July 2021).
Hate Crime: A Consultation Paper (2020) Law Commission Consultation Paper No 250; Intimate Image Abuse: A Consultation Paper (2021) Law Commission Consultation Paper No 253.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, paras 5.184-5.188.
This is more relevant to the non-criminal law response to online harms, as acknowledged in the Government response to the Online Harms White Paper: “Devolution” paras 6.15 to 17, available at: https://www.gov.uk/govemment/consultations/online-harms-white-paper/outcome/online-harms-white-paper-full-government-response (last visited 13 July 2021). Justice and policing are transferred matters on which the Northern Ireland Assembly has full legislative powers, by virtue of The Northern Ireland Act 1998 (Amendment of Schedule 3) Order 2010. See also: Guidance, Devolution settlement: Northern Ireland, 20 February 2013, updated 23 September 2019. Available at: https://www.gov.uk/guidance/devolution-settlement-northern-ireland (last visited 13 July 2021).
Hate Crime Legislation in Northern Ireland: Independent Review (November 2020), available at: https://www.justice-ni.gov.uk/sites/default/files/publications/justice/hate-crime-review.pdf (last visited 13 July 2021).
Department of Justice Northern Ireland, Consultation Response.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, paras 4.62.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, paras 4.63 to 66. (Citations omitted).
Crime and Disorder Act 1998, s 37(1).
Intimate Image Abuse: A Consultation Paper (2021), Law Com Consultation Paper No 253, para 1.75.
For example, Chapter 6, paras 6.112 and following.
Crown Prosecution Service, “Youth Offenders”, updated 28 April 2020, available at: https://www.cps.qov.uk/leqal-quidance/vouth-offenders (last visited 13 July 2021).
Crown Prosecution Service, “Social Media - Guidelines on prosecuting cases involving communications sent via social media” revised 21 August 2018, available at: https://www.cps.gov.uk/legal-guidance/social-media-guidelines-prosecuting-cases-involving-communications-sent-social-media (last visited 13 July 2021).
Sentencing Council, “Sentencing Children and Young People”, effective from 1 June 2017, available at: https://www.sentencingcouncil.org.uk/overarching-guides/magistrates-court/item/sentencing-children-and-young-people/ (last visited 13 July 2021).
Online Harms White Paper: Full UK government response to the consultation, “Interim measures from the government” para 2.46 and following; “Role of the government in media literacy” paras 5.23 to 28: https://www.gov.uk/government/consultations/online-harms-white-paper/outcome/online-harms-white-paper-full-government-response (last visited 13 July 2021).
See Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, at para 1.50.
A draft Online Safety Bill was published on 12 May 2021, available at:
https://www.gov.uk/government/publications/draft-online-safety-bill (last visited 13 July 2021).
Redmond-Bate v DPP [1999] EWHC Admin 733 [20]. It is worth also recalling the words of the European Court of Human Rights in Handyside v UK [1990] ECHR 32, 49, that the right to freedom of expression extends to those communications “that offend, shock or disturb the State or any sector of the population. Such are the demands of that pluralism, tolerance and broadmindedness without which there is no ‘democratic society’.”
Parliamentary Liaison and Investigation Team, Consultation Response.
National Police Chiefs’ Council, Consultation Response.
Criminal Bar Association, Consultation Response.
Magistrates Association, Consultation Response.
K Barker & O Jurasz, Consultation Response.
Refuge, Consultation Response.
Stonewall, Consultation Response.
ARTICLE 19, Consultation Response.
English PEN, Consultation Response.
Free Speech Union, Consultation Response.
It appears that this is a reference to the Post Office (Amendment) Act 1935. Lord Templemore is recorded in
Hansard: “.I come now to what is probably the most important clause of the whole Bill, Clause 10. Subsection (1) of that clause extends the penalties imposed by Section 67 of the Post Office Act, 1908, for obstructing officers of the Post Office to other forms of molestation and is designed to give the Post Office staff protection in cases where, for example, people have indulged in improper or obscene language over the telephone to female telephonists.” Hansard (HL) 19 March 1935, vol 96, col 165.
An either-way offence is one that can be tried either in the magistrates’ courts or in the Crown Court. If the
magistrates decide their sentencing powers are sufficient to deal with the offence, the accused may choose to have it dealt with summarily in the magistrates' court or on indictment (trial by jury) in the Crown Court.
An offence triable only in a magistrates’ court.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 5.1.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 5.2.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 4.121.
Association of Police and Crime Commissioners, Consultation Response.
Magistrates Association, Consultation Response.
Carnegie UK Trust, Consultation Response.
J Neller, Consultation Response.
English PEN, Consultation Response.
Crown Prosecution Service, Consultation Response.
A Gillespie, Consultation Response.
Bar Council, Consultation Response.
Parliamentary Liaison and Investigation Team, Metropolitan Police, Consultation Response.
Community Security Trust, Consultation Response.
E Tiarks & M Oswald, Consultation Response.
Criminal Bar Association, Consultation Response.
Stonewall, Consultation Response.
Magistrates Association, Consultation Response.
T Keren-Paz, Consultation Response.
WhatsApp Revenue and Usage Statistics (2020), available at:
https://www.businessofapps.com/data/whatsapp-statistics/ (last visited 13 July 2021).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No
248, consultation question 4.
A Gillespie, Consultation Response.
J Rowbottom, Consultation Response, para 7.
Criminal Law Solicitors’ Association, Consultation Response.
Parliamentary Liaison and Investigation Team, Metropolitan Police, Consultation Response.
Community Security Trust, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Stonewall, Consultation Response.
Refuge, Consultation Response.
A Gillespie, Consultation Response. Respectfully, we do not agree that in every case where actual harm
occurs that likely harm would not need to be proven. We deal with this further at para 2.107 below.
Magistrates Association, Consultation Response.
Law Society of England and Wales, Consultation Response.
This is a view fortified by the Divisional Court in Parkin v Norman [1983] QB 92, 100, in which the court held that, for the purposes of section 5 of the Public Order Act 1936, “likely to” indicated a higher degree of probability than “liable to”: “This is a penal measure and the courts must take care to see that the former expression is not treated as if it were the latter.” The offence in section 5 criminalised “threatening, abusive or insulting words or behaviour with intent to provoke a breach of the peace or whereby a breach of the peace is likely to be occasioned” (broadly the offence now contained in section 4 of the Public Order Act 1986).
Re H (Minors) [1995] UKHL 16, [1996] 1 AC 563, 584.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 3.
See para 2.110 above.
A Gillespie, Consultation Response.
Free Speech Union, Consultation Response.
Refuge, Consultation Response.
Bar Council, Consultation Response.
Judicial Security Committee, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Suzy Lamplugh Trust, Consultation Response.
The topic of “information subjects” is discussed further below.
E Tiarks & M Oswald, Consultation Response.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 6.
Fix the Glitch, Consultation Response.
Refuge, Consultation Response.
Women’s Aid, Consultation Response.
Sentencing Council, Overarching principles: domestic abuse, 24 May 2018. Available at:
https://www.sentencingcouncil.org.uk/overarching-guides/magistrates-court/item/domestic-abuse/ (last visited 13 July 2021).
Crown Prosecution Service, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Magistrates Association, Consultation Response.
English PEN, Consultation Response.
Free Speech Union, Consultation Response.
LGB Alliance, Consultation Response.
Sex Matters, Consultation Response.
See Murder, Manslaughter and Infanticide (2006), Law Commission Report No 304, paras 3.9-3.27. We discuss this in detail later in this section at para 2.186 onwards.
A draft Online Safety Bill was published on 12 May 2021, available at:
https://www.gov.uk/government/publications/draft-online-safetv-bill (last visited 13 July 2021).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No
248, consultation question 8.
R v G and another [2003] UKHL 50; [2004] 1 AC 1034.
T Keren-Paz, Consultation Response.
Young Epilepsy, Consultation Response.
A Gillespie, Consultation Response.
Refuge, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Crown Prosecution Service, Consultation Response.
End Violence Against Women Coalition, Consultation Response.
Magistrates Association, Consultation Response.
Criminal Bar Association, Consultation Response.
E Tiarks & M Oswald, Consultation Response.
Justices’ Legal Advisers and Court Officers’ Service, Consultation Response.
Kingsley Napley, Consultation Response.
Free Speech Union, Consultation Response.
ARTICLE 19, Consultation Response.
English PEN, Consultation Response.
See Murder, Manslaughter and Infanticide (2006), Law Commission Report No 304, para 3.27, and cited in Blackstone’s Criminal Practice at B1.15.
Murder, Manslaughter and Infanticide (2006), Law Commission Report No 304, para 3.27, and cited in Blackstone’s Criminal Practice at B1.15.
See Blackstone’s Criminal Practice at B1.14.
R v Moloney [1983] AC 905, 913E-F (Lord Hailsham of St Marylebone LC), affirmed by Lord Scarman in R v Hancock [1985] UKHL 9; [1986] 1 AC 455, 472B.
R v Hancock [1985] UKHL 9; [1986] 1 AC 455, 473F (Lord Scarman).
R v Moloney [1983] AC 905, 929H (Lord Bridge of Harwich), cited in R v Hancock [1985] UKHL 9; [1986] 1 AC 455, 472g and again referred to at 473d.
R v Nedrick [1986] 3 All ER 1; [1986] 1 WLR 1025.
R v Woollin [1999] 1 AC 82.
R v Nedrick [1986] 3 All ER 1; [1986] 1 WLR 1025, 1027-1028 (Lord Lane CJ).
R v Woollin [1999] 1 AC 82, 95F (Lord Steyn).
Indeed, we make a recommendation in Chapter 6: that purpose (of obtaining sexual gratification) should constitute one of the fault elements for the recommended offence of cyberflashing.
Chandler v DPP [1964] AC 763.
Hansard, HC Public Bill Committee, 6th Sitting, 3 July 2007, col.211.
See Blackstone’s Criminal Practice A5.12.
R v Moloney [1983] AC 905, 913E-F (Lord Hailsham of St Marylebone LC).
Of course, the excuse does still have to be reasonable: it is not enough that it is simply “an” excuse. It is certainly open to a court to find that there was no reasonable excuse for sending, say, Nazi propaganda dressed up as satire.
R v AY [2010] EWCA Crim 762.
R v AY [2010] EWCA Crim 762 at [25].
See our discussion of Peringek v Switzerland (2016) 63 EHRR 6 (App No 27510/08) in Chapter 2 of our Consultation Paper.
Sivier v Riley [2021] EWCA Civ 713, [20] (Warby LJ).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No
248, consultation question 11.
Health Protection (Coronavirus, Restrictions) Regulations (England) 2020.
For example: BBC News, 11 January 2021, “Covid: women fined for going for a walk receive police apology: available at: https://www.bbc.co.uk/news/uk-england-derbyshire-55625062 (last visited 13 July 2021). See also Joint Committee on Human Rights, “The Government response to covid-19: fixed penalty notices” 27 April 2021, available at: https://committees.parliament.uk/publications/5621/documents/55581/default/ (last visited 13 July 2021).
Justices’ Legal Advisers and Court Officers’ Service, Consultation Response.
Fair Cop, Consultation Response.
A Gillespie, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
English PEN, Consultation Response.
Magistrates Association, Consultation Response.
Bar Council, Consultation Response.
Kingsley Napley, Consultation Response.
Judicial Security Committee, Consultation Response.
Mermaids, Consultation Response.
Hacked Off, Consultation Response.
Fair Cop, Consultation Response.
J Neller, Consultation Response.
English PEN, Consultation Response.
Law Society of England and Wales, Consultation Response.
ARTICLE 19, Consultation Response.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 5.175.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No
248, paras 5.63 to 64.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 5.212 to 214.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 16.
Crown Prosecution Service, Consultation Response.
ARTICLE 19, Consultation Response.
For example, Refuge, Samaritans, Epilepsy Society, Community Security Trust, Consultation Responses.
A Gillespie, Community Security Trust, Consultation Responses.
See for example: A Mistlin, “Man who sent antisemitic tweets on holiday avoids UK prosecution” The Guardian, 23 February 2021. Available at: https://www.theguardian.com/news/2021/feb/23/man-who-sent-antisemitic-tweets-on-holiday-avoids-uk-prosecution (last visited 13 July 2021). See also L Harpin, “Police to drop Wiley antisemitism probe after learning he was abroad at time of alleged offences” Jewish Chronicle, 25 September 2020. Available at: https://www.thejc.com/news/uk/police-drop-wiley-antisemitism-probe-after-learning-he-was-abroad-at-time-of-alleged-offences-1.507042 (last visited 13 July 2021).
Domestic Abuse Act 2021, s 72 to 74, sch 3.
See for example: Domestic Abuse Act 2021, ss 72 to 74, sch 3; Sexual Offences Act 2003, s 72, sch 2; Suppression of Terrorism Act 1978, s 4, sch 1; Terrorism Act 2000, s 59, 62, 63; Terrorism Act 2006, s 17; Bribery Act 2010, s 12; Female Genital Mutilation Act 2003, s 4; Serious Crime Act 2015, s 70; Criminal Justice Act 1993, ss 1 to 3.
For example, Domestic Abuse Act 2021, s 72(1)(b).
Protection from Harassment Act 1997, ss 4 and 4A; Domestic Abuse Act 2021, sch 3.
Serious Crime Act 2015, s 76; Domestic Abuse Act 2021, sch 3.
Domestic Abuse Bill 2020, Extraterritorial jurisdiction factsheet, updated 18 May 2021. Available at:
https://www.qov.uk/qovemment/publications/domestic-abuse-bill-2020-factsheets/extraterritorial-jurisdiction-factsheet (last visited 13 July 2021).
Domestic Abuse Bill 2020, Extraterritorial jurisdiction factsheet, updated 18 May 2021. Available at:
https://www.qov.uk/qovernment/publications/domestic-abuse-bill-2020-factsheets/extraterritorial-jurisdiction-factsheet (last visited 13 July 2021).
Crown Prosecution Service, Jurisdiction, updated 2 September 2020. Available at: https://www.cps.qov.uk/leqal-quidance/jurisdiction (last visited 13 July 2021).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 5.212 to 214.
The current offence is found in the Fraud Act 2006, s 2.
Online Harms White Paper: Full UK government response to the consultation, “Disinformation and misinformation” at paras 2.75 and following, available at:
https://www.gov.uk/govemment/consultations/online-harms-white-paper/outcome/online-harms-white-paper-full-government-response (last visited 13 July 2021). We also differentiated “disinformation” and “misinformation” in our consultation paper at paragraph 6.43: Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248.
As we discussed in Example 2 of Chapter 6 of the consultation paper: Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.20.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, Chapter 3, paras 6.12, 6.36.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 19.
Association of Police and Crime Commissioners, Consultation Response.
S Rowe, Consultation Response.
English PEN, Consultation Response.
Alan Turing Institute, Consultation Response.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 2.20 to 59.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 3.125 to 3.139.
English PEN, Consultation Response.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.28.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.39 and following.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.41.
Ministry of Justice and Cabinet Office, Advice on introducing or amending criminal offences. Available at https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment data/file/481126/cr eating-new-criminal-offences.pdf (last visited 13 July 2021); Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.42.
Online Harms White Paper: Full government response to the consultation, “Disinformation and misinformation” 2.75ff, available at: https://www.gov.uk/government/consultations/online-harms-white-paper/outcome/online-harms-white-paper-full-government-response (last visited 13 July 2021).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.47 to 6.61.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 20.
Crown Prosecution Service, Consultation Response.
Fix the Glitch, Consultation Response.
Community Security Trust, Consultation Response.
English PEN, Consultation Response.
A Gillespie, Consultation Response.
Full Fact, Consultation Response.
Justices’ Legal Advisers and Court Officers’ Service, Consultation Response.
Community Security Trust, Consultation Response.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.45 to 6.46.
Hate Crime Laws: a Consultation Paper (2020) Law Commission Consultation Paper No 250, Chapter 18.
We note that the offences of stirring up hatred are not limited to stirring up racial hatred.
Hate Crime Laws: a Consultation Paper (2020) Law Commission Consultation Paper No 250, paras 18.142 to 152.
Hate Crime Laws: a Consultation Paper (2020) Law Commission Consultation Paper No 250, para 18.144. “The "blood libel” is antisemitic trope that Jews murder children in order to use their blood in religious rituals. Originating in Medieval England, where it was responsible for several massacres, the trope continues to circulate today.The Protocols were a hoax text purporting to document a Jewish conspiracy to control the world by controlling the press and finance. The Protocols continue to be cited by modern anti-Semites and conspiracy theorists.”
Magistrates Association, National AIDS Trust, Consultation Responses.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 4.104 to 4.132.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 4.113. Further, at paras 6.47 to 6.61 we acknowledge that in the context of knowingly false communications, there is a justification for including physical harm.
We also set out that, where appropriate, harms with a more diffuse causal connection to communications may well be better captured by other alternative offences that better reflect the nature and gravity of the harmful conduct, including offences against the person or sexual offences: Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 4.131.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 3.125 to 130.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 3.159 to 3.169.
National AIDS Trust, S Rowe, Demos, Carnegie UK Trust, Full Fact, Consultation Responses.
For example: S Rowe, A Gillespie, Consultation Responses.
The formulation “ought to have known” is also variously referred to as “wilful blindness”. We note the discussion of this in our report A Criminal Code for England and Wales (1989) Law Com No 177, vol 2, para 8.10. Even where a fault element requires intention, evidence of wilful blindness, or “shutting one’s eyes to the truth” may be used to infer knowledge or a belief: G Williams, Glanville Williams Textbook of Criminal Law (4th ed), para 6-017.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.41 to 6.46.
Big Brother Watch, Consultation Response.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.45.
See paras 2.59 to 2.60, 2.80.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 21.
Criminal Law Solicitors’ Association, Consultation Response.
S Rowe, Consultation Response.
LGB Alliance, Consultation Response.
English PEN, Consultation Response.
Refuge, Consultation Response.
See eg paras 1.15 to 1.17 and 2.180 to 2.183.
Communications Act 2003, s 127(3).
Justices’ Legal Advisers and Court Officers’ Service, Consultation Response.
Magistrates’ Courts Act 1980, s 127(1).
Criminal Justice and Courts Act 2015, s 51.
Criminal Justice and Courts Act 2015, Explanatory Notes s 51, available at:
https://www.legislation.gov.Uk/ukpga/2015/2/notes/division/3/3/2/1 (last visited 29 April 2021).
Criminal Justice and Courts Bill Fact Sheet: section 127 of the Communications Act 2003, available at: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment data/file/363924/fa ct-sheet-s127-comms-act.pdf (last visited 13 July 2021).
Criminal Justice and Courts Bill Fact Sheet: section 127 of the Communications Act 2003, para 6.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.13.
Scottow v Director of Public Prosecutions [2020] EWHC 3421 (Admin); [2021] Crim LR 315, [32].
Domestic Abuse Act 2021, s 39.
Domestic Abuse Act 2021, s 68; Serious Crime Act 2015, s 76.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.15; s 128 “Notification of misuse of networks and services” Communications Act 2003.
Namely “annoyance, inconvenience or anxiety”: Communications Act 2003, s 128(5).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 17.
Protection from Harassment Act 1997 s 1, 2 and 7.
For example, S Thomas, N Djaba, Consultation Responses.
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248.
Justices’ Legal Advisers and Court Officers’ Service, Consultation Response.
National Police Chiefs’ Council, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
ARTICLE 19, Consultation Response.
[2020] EWHC 3421 (Admin); [2021] Crim LR 315.
Free Speech Union, Consultation Response.
Magistrates Association, Consultation Response.
This may include traditional fixed line phone and mobile telephone networks, but also Voice-over-InternetProtocol (VoIP) or similar broadband-based phone services.
For example, Criminal Law Act 1977, s 51 “Bomb hoaxes”; Anti-terrorism, Crime and Security Act 2001,
s 114 “Hoaxes involving noxious substances or things”.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.11 to 6.17.
Communications Act 2003, s 127(3).
For example, Offences Against the Person Act 1861, s 16 “Threats to kill”; bomb hoaxes are covered by the Criminal Law Act 1977, s 51. We described the existing position regarding threats as lacking coherence and amounting to a “patchwork of different statutes": Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, Chapter 7.
Reform of Offences Against the Person (2015) Law Com No 361, para 8.18.
Intimate Image Abuse: A consultation paper (2021) Law Commission Consultation Paper No 253, paras
12.1 and following.
Chambers v Director of Public Prosecutions [2012] EWHC 2157 (Admin); [2013] 1 WLR 1833.
Chambers v Director of Public Prosecutions [2012] EWHC 2157 (Admin); [2013] 1 WLR 1833, paras [26] to [34].
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 5.207, citing: BBC News, YouTuber jailed for Birmingham hospital bomb threat (18 June 2020), https://www.bbc.co.uk/news/uk-england-birmingham-53092117 (last visited 13 July 2021).
UK Government National Action Plan for the Safety of Journalists, 9 March 2021, available at:
https://www.gov.uk/government/publications/national-action-plan-for-the-safety-of-joumalists/national-action-plan-for-the-safety-of-journalists (last visited 13 July 2021).
R Syval, “UK launches action plan to prevent harassment and abuse of journalists” The Guardian 9 March 2021, available at: https://www.theguardian.com/media/2021/mar/09/uk-launches-action-plan-prevent-harassment-abuse-journalists (last visited 13 July 2021).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 5.210.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 15.
A Gillespie, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Refuge, Consultation Response.
English PEN, Consultation Response.
Suzy Lamplugh Trust, Consultation Response
Fix the Glitch, Consultation Response.
Z Alhaboby, Consultation Response.
Judicial Security Committee, Consultation Response.
LGB Alliance, Consultation Response.
Refuge, Consultation Response.
Women’s Aid, Consultation Response.
Antisemitism Policy Trust, Consultation Response.
Suzy Lamplugh Trust, Consultation Response.
ARTICLE 19, Consultation Response.
Criminal Bar Association, Consultation Response.
Fair Cop, Consultation Response.
Free Speech Union, Consultation Response.
Crown Prosecution Service, Consultation Response.
This can be contrasted with the maximum penalty of 10 years’ imprisonment for the offence of threats to kill: Offences Against the Person Act 1861, s 16.
Chambers v Director of Public Prosecutions [2012] EWHC 2157 (Admin); [2013] 1 WLR 1833.
Protection from Harassment Act 1997, s 1.
For example: Domestic Abuse Act 2021, s 39, 68; Serious Crime Act 2015, s 76.
Intimate Image Abuse: A consultation paper (2021) Law Commission Consultation Paper No 253, Chapter 12.
Intimate Image Abuse: A consultation paper (2021) Law Commission Consultation Paper No 253, paras 12.84 to 12.108.
Intimate Image Abuse: A consultation paper (2021) Law Commission Consultation Paper No 253, paras 12.113 to 114.
Intimate Image Abuse: A consultation paper (2021) Law Commission Consultation Paper No 253, paras 12.144 to 145.
See the discussion in Intimate Image Abuse: A consultation paper (2021) Law Commission Consultation Paper No 253, Chapter 14.
Reform of Offences against the Person (2015) Law Com No 361, para 8.18.
Misconduct in Public Office (2020) Law Com No 397, paras 6.88 to 6.89.
As discussed in R v Ireland [1988] AC 147 this will incorporate “recognisable psychiatric illness”.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 5.200.
A Gillespie, Consultation Response.
Intimate Image Abuse: A consultation paper (2021) Law Commission Consultation Paper No 253, para 12.138.
For example, Malicious Communications Act 1988, s 1(2); Theft Act 1968, s 21(1)(a) and (b).
For further discussion, see for example “Defences involving other excuses and justifications” Blackstone’s Criminal Practice para A3.34 and following.
Offences Against the Person Act 1861, s 16 “Threats to kill”.
Reform of Offences against the Person (2015) Law Com No 361, para 8.18.
Epilepsy Society, Zach’s Law: protecting people with epilepsy from online harms (July 2020). See also S Elvin, Trolls attack boy with epilepsy, 8, by sending him hundreds of flashing images (29 May 2020), https://metro.co.uk/2020/05/29/trolls-attack-boy-epilepsy-8-sending-hundreds-flashing-images-12777199/ (last visited 13 July 2021).
Epilepsy Society, Consultation Response, p 5.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 4.123-131.
See for example, Epilepsy, “Symptoms”, NHS (last reviewed 18 September 2020), available at: https://www.nhs.uk/conditions/epilepsy/symptoms/ (last visited 13 July 2021).
Reform of Offences Against the Person (2015) Law Com No 361, paras 4.13 to 33.
Hacked Off, Consultation Response.
Gina Miller, Consultation Response.
Media Lawyers’ Association, Consultation Response.
News Media Association, DMG Media, Consultation Responses.
Criminal Bar Association, Consultation Response.
Crown Prosecution Service, Consultation Response.
IMPRESS, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Fix the Glitch, Consultation Response.
National Secular Society, Consultation Response.
Free Speech Union, Consultation Response.
Mermaids, Consultation Response.
Crown Prosecution Service, Consultation Response.
Bar Council, Consultation Response.
News Media Association, Consultation Response.
Media Lawyers’ Association, Consultation Response.
Media Lawyers’ Association, Consultation Response.
Community Security Trust, Consultation Response.
See also, “News website accused of antisemitism joins press regulator” The JC, 21 March 2021, available at: https://www.thejc.com/news/uk/news-website-accused-of-antisemitism-joins-press-regulator-1.513341 (last visited 13 July 2021).
Carnegie UK Trust, Consultation Response.
In law, “intention” is not the same as “purpose”. A person can be (but does not have to be) deemed to have intended those outcomes which he or she foresaw as a virtually certain consequence of his or her actions. Purpose, by contrast, is the desired outcome. So a person who blows up an aircraft full of people hoping to collect on the insurance money has, as his or her purpose, an insurance windfall, but clearly intends the death of those he or she kills. See also the discussion of intention in Chapter 2: .
Online Harms White Paper: Full government response to the consultation, 15 December 2020, “Journalism”, available at: https://www.gov.uk/govemment/consultations/online-harms-white-paper/outeome/online-harms-white-paper-full-govemment-response (last visited 13 July 2021).
Draft Online Safety Bill, published 12 May 2021, available at:
https://www.gov.uk/govemment/publications/draft-online-safety-bill (last visited 13 July 2021).
We note that should a different approach be taken in relation to the Online Harms framework, existing definitions of journalism could be used to provide effective levels of protection, for example the protections afforded when things are done “for the purposes of journalism” in the Investigatory Powers Act 2016, s 264.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.72; Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, para 3.78
Amnesty International, UK: Twitter still failing women over online violence and abuse - new analysis (20 September 2020), available at: https://www.amnesty.org.uk/press-releases/twitter-still-failing-women-over-online-violence-and-abuse-new-analysis (last visited 13 July 2021); see also Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.74.
See E Vogels, “The State of Online Harassment” Pew Research Centre, January 13 2021, available at: https://www.pewresearch.org/internet/2021/01/13/the-state-of-online-harassment/ (last visited 13 July 2021): “Overall, men are somewhat more likely than women to say they have experienced any form of harassment online (43% vs. 38%), but similar shares of men and women have faced more severe forms of this kind of abuse. There are also differences across individual types of online harassment in the types of negative incidents they have personally encountered online. Some 35% of men say they have been called an offensive name versus 26% of women and being physically threatened online is more common occurrence for men rather than women (16% vs. 11%). Women, on the other hand, are more likely than men to report having been sexually harassed online (16% vs. 5%) or stalked (13% vs. 9%). Young women are particularly likely to have experienced sexual harassment online. Fully 33% of women under 35 say they have been sexually harassed online, while 11% of men under 35 say the same. Lesbian, gay or bisexual adults are particularly likely to face harassment online. Roughly seven-in-ten have encountered any harassment online and fully 51% have been targeted for more severe forms of online abuse.”.
Fix the Glitch UK and End Violence Against Women Coalition, “The Ripple Effect: COVID-19 and the Epidemic of Online Abuse”, available at: https://www.endviolenceagainstwomen.org.uk/wp-content/uploads/Glitch-and-EVAW-The-Ripple-Effect-Online-abuse-during-COVID-19-Sept-2020.pdf (last visited 13 July 2021).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 3.36. See also the discussion by Warby J in Hourani v Thompson [2017] EWHC 432 (QB) at [131]-[135], a civil case concerning an orchestrated “campaign” of harassment, where a defendant’s argument that their single instance of conduct was insufficient to amount to a “course of conduct” was rejected. It was held that the conduct of the co-defendants could be attributed to the defendant by virtue of s 7(3A).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 3.63 to 66.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.78 to 84. We note that the reference to the “explicit protection for political speech” is to the “contribution to a matter of public interest” aspect of the without reasonable excuse element of the offence discussed in both the consultation paper and Chapter 2 above.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.86.
The following paragraphs (5.21 to 5.29) are only very slightly modified from the Consultation Paper: Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.87 to 99.
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, paras 8.46 to 8.47.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.87ff.
Crown Prosecution Service, Guidelines on prosecuting cases involving communications sent via social media (last revised 21 August 2018) available at https://www.cps.gov.uk/legal-guidance/social-media-guidelines-prosecuting-cases-involving-communications-sent-social-media (last visited 13 July 2021).
An either-way offence is one that can be tried either in the magistrates’ courts or in the Crown Court. If the magistrates decide their sentencing powers are sufficient to deal with the offence, the accused may choose to have it dealt with summarily in the magistrates' court or on indictment (trial by jury) in the Crown Court.
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, paras 8.206 to 8.207.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 22.
Magistrates Association, Consultation Response.
Alan Turing Institute, Consultation Response.
Community Security Trust, Consultation Response.
Refuge, Consultation Response.
Fix the Glitch, Consultation Response.
Stonewall, Consultation Response.
Epilepsy Society, Consultation Response.
Suzy Lamplugh Trust, Consultation Response.
A Gillespie, Consultation Response.
Bar Council of England and Wales, Consultation Response.
Criminal Bar Association, Consultation Response.
Sex Matters, Consultation Response.
Adam Smith Institute, Consultation Response.
Criminal Law Solicitors’ Association, Consultation Response.
Fair Cop, Consultation Response.
ARTICLE 19, Consultation Response.
Free Speech Union, Consultation Response.
English PEN, Consultation Response.
Crown Prosecution Service, Consultation Response.
Carnegie UK Trust, Consultation Response.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.87 to 98.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.87.
Protection from Harassment Act 1997, s 7(3A).
There is a separate, but related question as to whether Oscar’s post encouraging the “pile-on” could itself be an instance of conduct for the purpose of the PHA 1997, such that if only one other person responds to Oscar’s call to “teach Ruby a lesson” there would be a “course of conduct”. In our view, it would be possible that though not directly contacting Ruby, Oscar’s message could be included in a “course of conduct”. Further, as we set out in the scoping report and consultation paper, and as our consultees pointed out in their responses on this matter, most instances of “pile-on” harassment involve a very large number of communications. To that end, the culpable instances of inciting or encouraging “pile-on” harassment will likely involve many more than two individual communications.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.89.
Protection from Harassment Act 1997, s 4.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.90 to 6.98.
Crown Prosecution Service, Social Media: Guidelines on prosecuting cases involving communications sent via social media (last revised 21 August 2018), para 15. Available at: https://www.cps.gov.uk/legal-guidance/social-media-guidelines-prosecuting-cases-involving-communications-sent-social-media (last visited 13 July 2021).
For example, Criminal Bar Association, Bar Council of England and Wales, Criminal Law Solicitors’ Association, English PEN, Consultation Responses. Further, where a person repeatedly attempts, unsuccessfully, to incite or encourage a “pile-on” against a person, the ordinary provisions of the PHA 1997 may well apply, as those repeated unsuccessful attempts may constitute a “course of conduct” that “amounts to harassment”.
Online Harms White Paper: Full UK government response to the consultation, paras 2.19 to 2.27, available at: https://www.gov.uk/govemment/consultations/online-harms-white-paper/outeome/online-harms-white-paper-full-govemment-response (Last visited 13 July 2021).
Online Harms White Paper: Full UK government response to the consultation, paras 2.28 to 2.34, available at: https://www.gov.uk/govemment/consultations/online-harms-white-paper/outeome/online-harms-white-paper-full-govemment-response (Last visited 13 July 2021).
Protection from Harassment Act 1997, s 1.
Protection from Harassment Act 1997, s 7.
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, para 8.207.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.100 to 6.104.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 23.
Fix the Glitch, Consultation Response.
K Barker & O Jurasz, Consultation Response.
T Keren-Paz, Consultation Response.
Criminal Bar Association, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Magistrates Association, Consultation Response.
Alan Turing Institute, Consultation Response.
Bar Council of England and Wales, Consultation Response.
Crown Prosecution Service, Consultation Response. A similar point was made regarding the investigation of any potential new offence by the Association of Police and Crime Commissioners in their Consultation Response.
Refuge, Consultation Response.
Kingsley Napley, Consultation Response.
ARTICLE 19, Consultation Response.
Sex Matters, Consultation Response.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.104.
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, Chapter 12.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.152.
J Ilan, ‘Digital Street Culture Decoded: Why Criminalizing Drill Music Is Street Illiterate and Counterproductive’ (2020) British Journal of Criminology, No 4, pp 994 to 1013.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.153.
See eg Criminal Behaviour Orders under the Anti-Social Behaviour, Crime and Policing Act 2014.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.154 to 158.
JUSTICE, “Tackling Racial Injustice: Children and the Youth Justice System” (2021), 33-34. Available at: https://justice.org.uk/wp-content/uploads/flipbook/46/book.html (last visited 13 July 2021).
JUSTICE, “Tackling Racial Injustice: Children and the Youth Justice System” (2021), 40; citing S Swan, “Drill and rap music on trial” BBC News, 13 January 2021, available at: https://www.bbc.co.uk/news/uk-55617706 (last visited 13 July 2021).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.153; JUSTICE, “Tackling Racial Injustice: Children and the Youth Justice System” (2021), 49.
JUSTICE, “Tackling Racial Injustice: Children and the Youth Justice System” (2021), 101; M Keenan, Youth Justice Centre, “JUSTICE report: report finds misunderstanding of Drill music is leading to unfair convictions” 11 March 2021, available at: https://yjlc.uk/resources/legal-updates/justice-report-report-finds-misunderstanding-drill-music-leading-unfair (last visited 13 July 2021).
Derek Chauvin, a former police officer, was convicted on April 21, 2021 of charges of second-degree murder, third-degree murder and manslaughter for the death of Mr Floyd which occurred on 25 May 2020. See BBC News: “George Floyd: Jury finds Derek Chauvin guilty of murder” available at: https://www.bbc.co.uk/news/world-us-canada-56818766 (last visited 13 July 2021).
See, for example, A Hern, Twitter hides Donald Trump tweet for 'glorifying violence' The Guardian (29 May 2020), available at https://www.theguardian.com/technology/2020/may/29/twitter-hides-donald-trump-tweet-glorifying-violence (last visited 13 July 2021).
Twitter, “Permanent suspension of @realDonaldTrump” 8 January 2021, available at:
https://blog.twitter.com/en us/topics/company/2020/suspension.html (last visited 13 July 2021).
Twitter, Glorification of violence policy March 2019: “We define glorification to include praising, celebrating, or condoning statements, such as “I’m glad this happened”, “This person is my hero”, “I wish more people did things like this”, or “I hope this inspires others to act”.” Available at: https://help.twitter.com/en/rules-and-policies/glorification-of-violence (last visited 13 July 2021).
See the discussion in Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, para 12.88 and Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.159 to 165.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.161.
Terrorism Act 2000, s 1 and s 2.
See Reform of Offences Against the Person (2015) Law Com No 361. The government has not yet responded to the report.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.164 to 165.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.166 to 174.
See, for example, Index on Censorship, ‘A guide to the legal framework impacting on artistic freedom of expression’, available at https://www.indexoncensorship.org/wp-content/uploads/2015/07/Counter-Terrorism 210715.pdf (last visited 13 July 2021).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 27.
Community Security Trust, Antisemitism Policy Trust, Consultation Responses.
LGB Alliance, LGBT Fed, Consultation Responses.
Fix the Glitch, Consultation Response.
National Police Chiefs’ Council, Consultation Response.
Kingsley Napley, Consultation Response.
J Neller, Consultation Response. We note that such an amendment would likely have wide-ranging effects and be outside the scope of this project.
Magistrates Association, Consultation Response.
ARTICLE 19, Consultation Response.
Criminal Bar Association, Consultation Response.
English PEN, Consultation Response.
Crown Prosecution Service, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Bar Council of England and Wales, Consultation Response.
Commission for Countering Extremism, “Operating with Impunity: Hateful extremism: the need for a legal
framework” (February 2021), available at:
CE Operating with Impunity Accessible.pdf (Last visited 13 July 2021), especially paras 6.1 to 6.39.
Contained in our Report: Reform of Offences Against the Person (2015) Law Com No 361.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.173.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.200 to 206.
[2018] EWCA Crim 560; [2018] Crim LR 847.
For discussion of the medical exception, see P Lewis, The Medical Exception (2012) 65 Current Legal Problems 355.
See The Guardian, 21 March 2019, available at: https://www.theguardian.com/uk-
news/2019/mar/21/tattooist-dr-evil-jailed-for-performing-ear-and-nipple-removals (last visited 13 July 2021).
See Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, Chapter
12.
Serious Crime Act 2007, s 58(3). “Anticipated or reference offence” means the offence that is encouraged or assisted.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, consultation question 30.
Justices’ Legal Advisers and Court Officers’ Service, Consultation Response.
ARTICLE 19, Consultation Response.
Criminal Bar Association, Consultation Response.
See paras 7.17 to 7.19 in Chapter 7 for a discussion of the possible application of s 18 of the Offences Against the Person Act 1861 where a person harms themselves.
[2018] EWCA Crim 560; [2018] Crim LR 847.
See Chapter 7: .
See, for example, Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, para 4.82.
C McGlynn and K Johnson, Consultation Response, p 5, citing A Marcotte, A Gesselman, H Fisher & J Garcia, “Women’s and Men’s Reactions to Receiving Unsolicited Genital Images from Men” (2020) The Journal of Sex Research; J Boulos “Cyber Flashing: ‘I froze when penis picture dropped on to my phone’” BBC News 26 April 2019, available at: https://www.bbc.co.uk/news/uk-48054893 (last visited 13 July 2021); A Gillespie, “Tackling Voyeurism: Is The Voyeurism (Offences) Act 2019 A Wasted Opportunity?” (2019) 82 Modern Law Review 1107.
The Association of Police and Crime Commissioners, Consultation Response.
S Gallagher, Consultation Response.
https://anqelou-centre.orq.uk/?paqe id=92 (last visited 13 July 2021).
The Angelou Centre, Consultation Response.
National Police Chiefs’ Council, Consultation Response.
Suzy Lamplugh Trust, Consultation Response.
S Gallagher, Consultation Response.
C McGlynn and K Johnson, Consultation Response, p 6.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, para 6.106.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, para 4.82.
National Police Chiefs’ Council, Consultation Response.
Fix the Glitch, Consultation Response.
C McGlynn and K Johnson, Consultation Response, p 3, citing J Ringrose, “Is there hidden sexual abuse going on in your school?” TES, 29 October 2020. Available at: https://www.tes.com/news/there-hidden-sexual-abuse-going-your-school (last visited 13 July 2021).
C McGlynn and K Johnson, Consultation Response, p 3.
See, for example, Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, para 3.156.
See Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, paras 3.9 to 3.12.
Crown Prosecution Service, Consultation Response.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, consultation question 24.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, paras 6.124 to 6.125. It is worth noting that an offence does not have to be contained in the Sexual Offences Act 2003 itself in order to be included in Schedules 3 and 5.
R v Alderton [2014] EWCA Crim 2204.
Bar Council of England & Wales, Consultation Response.
Crown Prosecution Service, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Angelou Centre, Consultation Response.
It is important to clarify that it is not up to the victim whether a Sexual Harm Prevention Order is made.
Suzy Lamplugh Trust, Consultation Response.
Magistrates Association, Consultation Response.
C McGlynn and K Johnson, Consultation Response, p 7.
End Violence Against Women Coalition, Consultation Response.
Big Brother Watch, Consultation Response.
Free Speech Union, Consultation Response.
Sexual Offences Act 2003, s 3.
As we set out in Intimate Image Abuse: A consultation paper (2021) Law Commission Consultation Paper No 253, paras 6.5 to 6.12, other similar offences capture photographs and video. Further, our consultation questions all specifically addressed images and video recordings. While we may refer at times simply to an “image”, this is simply for the sake of brevity, and is not intended to reflect a narrower scope.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, para 1.21.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, consultation question 25.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, paras 6.139 to 6.140.
Refuge, Consultation Response.
A Gillespie, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Magistrates Association, Consultation Response.
L Thompson, Consultation Response.
End Violence Against Women Coalition, Consultation Response.
C McGlynn and K Johnson, Consultation Response, p 9.
S Gallagher, Consultation Response.
T Keren-Paz, Consultation Response.
Crown Prosecution Service, Consultation Response.
Fix the Glitch, Consultation Response.
S Thomas, Consultation Response.
See, for example, the Office of National Statistics’ Crime Survey for England and Wales, “Sexual Offences
in England and Wales: year ending March 2017” available at:
https://www.ons.qov.uk/peoplepopulationandcommunity/crimeandjustice/articles/sexualoffencesinenqlandan dwales/yearendingmarch2017 (last visited 13 July 2021).
Hansard (HL), 19 May 2003, vol 648, col 556.
It is for this reason that we do not take forward the suggestion made by Professor Keren-Paz at 6.68, above,
to limit the offence to images of the sender or that the recipient reasonably believed were of the sender.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, paras 6.146 to 6.148.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, consultation question 26(1).
#NotYourPorn, Consultation Response.
C McGlynn and K Johnson, Consultation Response, p 8.
Angelou Centre, Consultation Response.
End Violence Against Women Coalition, Consultation Response.
Crown Prosecution Service, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
L Thompson, Consultation Response.
Criminal Bar Association, Consultation Response.
Bar Council of England and Wales, Consultation Response.
A Gillespie, Consultation Response.
Fair Cop, Consultation Response.
Kingsley Napley, Consultation Response.
Kingsley Napley, Consultation Response.
Kingsley Napley, Consultation Response.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, consultation question 26(2).
Justices’ Legal Advisers and Court Officers’ Service, Consultation Response.
A Gillespie, Consultation Response.
Angelou Centre, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Magistrates Association, Consultation Response.
Fair Cop, Consultation Response.
Criminal Law Solicitors’ Association, Consultation Response.
Criminal Bar Association, Consultation Response.
G Miller, Consultation Response.
Intimate Image Abuse (2021) Law Commission Consultation Paper No 253, para 8.20.
We note that criminal offences targeting defendants acting for this purpose already exist. For example, sections 12, 33 and 67 of the Sexual Offences Act 2003.
R v G [2003] UKHL 50. See also the discussions of recklessness in our reports - Criminal Law: A Criminal Code for England and Wales (1989) Law Com No 177, vol 1 and Legislating the Criminal Code: Offences Against the Person and General Principles (1993) Law Com No 218.
For example, campaigns and resources directed toward better education on consent: Personal, Social Health and Economic Association, Guidance on teaching about consent: https://www.pshe-association.org.uk/curriculum-and-resources/resources/guidance-teaching-about-consent-pshe-education-key; Schools Consent Project https://www.schoolsconsentproject.com/; Department for Education Guidance on Relationships and sex education (RSE) and health education, last updated 9 July 2020, available at: https://www.gov.uk/government/publications/relationships-education-relationships-and-sex-education-rse-and-health-education (links last visited 13 July 2021).
Crown Prosecution Service, Youth Offenders guidance, “Section 1 of the Voyeurism (Offences) Act 2019 -‘upskirting’” 28 April 2020, available at: https://www.cps.gov.uk/legal-guidance/youth-offenders (last visited 13 July 2021).
The threshold for exposure (for adult offenders) is met when either the victim was under 18, or the offender has been sentenced to a term of imprisonment, detained in a hospital, or made the subject of a community sentence of at least 12 months. Sexual Offences Act 2003, Schedule 3, regulation 33.
Sexual Offences Act 2003, Schedule 3, regulation 34A.
For example, various provisions of the Youth Justice and Criminal Evidence Act 1999 provide for: giving evidence behind a screen to shield the witness from the defendant (s 23), giving evidence via live link to allow evidence to be given from outside the courtroom (s 24), giving evidence in private (s 25), removal of wigs and gowns (s 26), and video-recorded evidence in chief (s 22A).
Intimate Image Abuse: A consultation paper (2021) Law Com Consultation Paper No 253, paras 14.68 to 84.
Report of the Advisory Group on the Law of Rape (1975) Cmnd 6352.
For example, s 122A of the Antisocial Behaviour, Crime and Policing Act 2014 (for victims of forced marriages), and s 4A of the Female Genital Mutilation Act 2003 (for victims of female genital mutilation).
Sexual Offences (Amendment) Act 1992, s 2(1)(da).
Youth Justice and Criminal Evidence Act 1999, s 17.
Youth Justice and Criminal Evidence Act 1999, sections 34 & 35 (prohibiting self-represented defendants in trials of sexual offences cross-examining complainants and child complainants and child witnesses); s 41 (restrictions on adducing evidence of or questioning the complainant about their sexual behaviour).
Youth Justice and Criminal Evidence Act 1999, s 62.
We note that the offence of suicide was abolished by the Suicide Act 1961.
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Com No 381, 12.91.
See, for example, N Shanahan and others, ‘Self-harm and social media: thematic analysis of images posted on three social media sites’ (2019) 9(2) BMJ Open, available at:
https://bmjopen.bmj.eom/content/9/2/e027006#ref-12 (last visited 13 July 2021). See also R Brown and others ‘#cutting: non-suicidal self-injury (NSSI) on Instagram’ (2018) 48(2) Psychological Medicine 337.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.178.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.178.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.180 to 181.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.182.
Government Serious and Organised Crime Strategy Paper, 27, available at:
https://assets.publishing.service.gov.uk/govemment/uploads/system/uploads/attachment data/file/752850/S OC-2018-web.pdf (last visited 13 July 2021). Whether this conduct provides a “paradigm” example of encouragement of self-harm is potentially open to debate. However, it demonstrates the potential for malign actors to manipulate people who are at risk and that self-harm content is not solely created and shared by people who are themselves vulnerable.
A charity “dedicated to empowering the safe and secure use of technology” whose services include the support helplines Report Harmful Content and the Revenge Porn Helpline: https://swgfl.org.uk/about/ (last visited 13 July 2021).
Abusive and Offensive Online Communications: A Scoping Report (2018) Law Commission Scoping Report Law Com No 381.
Inchoate Liability for Assisting and Encouraging Crime (2006) Law Com No 300, para 1.9.
Inchoate Liability for Assisting and Encouraging Crime (2006) Law Com No 300, para 1.9.
Serious Crime Act 2007 section 44 “intentionally encouraging or assisting an offence”, section 45 “encouraging or assisting an offence, believing it will be committed”, and section 46 “encouraging or assisting offences, believing one or more will be committed”.
Blackstone’s Criminal Practice 2021 Chapter A5 Inchoate Offences A5.4.
Blackstone’s Criminal Practice 2021 Chapter A5 Inchoate Offences A5.6.
Blackstone’s Criminal Practice 2021 Chapter A5 Inchoate Offences A5.32.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.185.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.185.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.185 to 6.189.
See our separate report, Reform of Offences against the Person (2015) Law Commission Report Law Com No 361.
Reform of Offences against the Person (2014) Law Commission Consultation Paper 217, para 2.122.
W Blackstone, Commentaries on the Laws of England (1st ed 1765-1769) vol 4, p 205.
Smith, Hogan & Ormerod’s Criminal Law (15th ed) Chapter 11 “Inchoate Crime” 11.4.4, 481 (citations in original).
[1998] 1 Cr App R (S) 273, [1997] Crim LR 897, CA. (A statutory offence of incitement, contrary to the
Misuse of Drugs Act 1971, s 19.
Invicta Plastics Ltd [1976] RTR 251, [1976] Crim LR 131, DC. (Indication that ‘Radatex’ may be used to detect police radar traps was incitement to an offence under s 1(1) of the Wireless Telegraphy Act 1949). Note that the licensing requirement was removed by SI 1989/123, and that no offence of ‘obtaining information’ is committed by the user of the apparatus: Knightsbridge Crown Court, ex parte Foot [1999] RTR 21.) Cf the reports in The Times, 6 Aug 1998 that a student was convicted of ‘inciting speeding offences’ by the sale of a ‘speed trap jammer’. In Parr-Moore [2002] EWCA Crim 1907, [2003] 1 Cr App R (S) 425, the court described the appellants’ publication of disclaimer on such a device serving only to illustrate their realization that the trade was illegal, at [3].
James and Ashford (1985) 82 Cr App R 226 at 232, [1986] Crim LR 118, CA, distinguishing Invicta Plastics
Ltd. See also Maxwell-King [2001] 2 Cr App R (S) 28, (2001) The Times, 2 Jan, CA: incitement to commit offence contrary to the Computer Misuse Act 1990, s 3, by supply of device to allow unauthorized access to satellite TV channels.
Blackstone’s Criminal Practice 2021 Chapter B1 Homicide and Related Offences, B1.145.
See s 2(1)(b) Suicide Act 1961 extracted above. The previous formulation is available at:
https://www.legislation.gov.uk/ukpga/Eliz2/9-10/60/1991-02-01 (last visited 13 July 2021) and referred to in A-G v Able [1984] 1 QB 795.
SWGfL & A Phippen, Consultation Response.
Facebook, Consultation Response.
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No
248, consultation question 28.
Samaritans, Consultation Response.
Refuge, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Magistrates Association, Consultation Response.
Big Brother Watch, Consultation Response.
For more information on the “public interest test” applied by prosecutors, see: Crown Prosecution Service, Suicide: Policy for Prosecutors in Respect of Cases of Encouraging or Assisting Suicide, February 2010, updated October 2014, available at: https://www.cps.gov.uk/legal-guidance/suicide-policy-prosecutors-respect-cases-encouraging-or-assisting-suicide (last visited 13 July 2021).
Criminal Bar Association, Consultation Response.
Confusingly, the term “reasonable excuse” is frequently used to refer to justifications or permissions. Ashworth and Horder, Principles of Criminal Law 7th ed 2013, 116.
This is considered in some detail in the Full government response to the consultation on the Online Harms White Paper, for example at paras 2.28 to 2.45, Available at:
https://www.gov.uk/govemment/consultations/online-harms-white-paper/outcome/online-harms-white-paper-full-government-response (last visited 13 July 2021).
Harmful Online Communications: A Consultation Paper (2020) Law Commission Consultation Paper No 248, consultation question 29.
Samaritans, Consultation Response.
Her Majesty’s Government, Consultation Response.
Criminal Bar Association, Consultation Response.
Association of Police and Crime Commissioners, Consultation Response.
Law Society of England and Wales, Consultation Response.
Bar Council of England and Wales, Consultation Response.
ARTICLE 19, Consultation Response.
Magistrates Association, Consultation Response.
R v Natasha Gordon [2018] EWCA Crim 1803.
SWGfL’s general response contained similar observations about the risks that attend “digital ghost stories”.
Samaritans, confidential “accompanying note” and “Rapid Review of Evidence”. These were provided
confidentially and any inclusion in a final published paper or report would be subject to Samaritans’ review and agreement.
Director of Public Prosecutions, “Suicide: Policy for Prosecutors in Respect of Cases of Encouraging or Assisting Suicide” February 2010, updated October 2014. Available at: https://www.cps.gov.uk/legal-guidance/suicide-policy-prosecutors-respect-cases-encouraging-or-assisting-suicide (last visited 13 July 2021).
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.195 to 6.198.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, paras 6.195 to 6.198. We note that Brazil recently extended their incitement to suicide offence to criminalise “inducement or instigation” of self-harm: https://www.loc.gov/law/foreign-news/article/brazil-new-law-criminalizes-instigation-to-self-harm/ (last visited 13 July 2021).
For a detailed overview of the Act, and our own proposals for reform, see Reform of Offences Against the Person (2015) Law Com Report No 361.
Miller [1954] 2 QB 282, [1954] 2 All ER 529.
Reform of Offences Against the Person (2015) Law Com Report No 361, para 2.19.
See R v Brown [1993] UKHL 19, [1994] 1 AC 212.
See also BBC News, “’Concerning’ rise in pre-teens self-injuring” (16 February 2021), available at: https://www.bbc.co.uk/news/uk-55730999 (last visited 13 July 2021).
Discussed at length in Re A (Children) (Conjoined Twins: Medical Treatment) (No 1) [2001] Fam 147, and in our separate Report: Defences of General Application (1977) Law Com Report No 83.
Harmful Online Communications: The Criminal Offences (2020) Law Commission Consultation Paper No 248, para 6.197.
For example, the Consultation Responses of Samaritans, SWGfL, B Thain, C Wilton-King, Association of
Police and Crime Commissioners all referred to the importance of accessible mental health resources.
Consultation Responses that noted this included those from the Crown Prosecution Service, A Gillespie, Association of Police and Crime Commissioners, National Police Chiefs’ Council, Bar Council of England and Wales, Criminal Bar Association.
Director of Public Prosecutions, “Suicide: Policy for Prosecutors in Respect of Cases of Encouraging or Assisting Suicide” February 2010, updated October 2014. Available at: https://www.cps.gov.uk/legal-guidance/suicide-policy-prosecutors-respect-cases-encouraging-or-assisting-suicide (last visited 13 July 2021).
Director of Public Prosecutions, “Suicide: Policy for Prosecutors in Respect of Cases of Encouraging or
Assisting Suicide” February 2010, updated October 2014, “Public interest factors tending against prosecution”. Available at: https://www.cps.gov.uk/legal-guidance/suicide-policy-prosecutors-respect-cases-encouraging-or-assisting-suicide (last visited 13 July 2021).
As for the Suicide Act 1961, we would not recommend that this must be the personal consent of the Director of Public Prosecutions. See Crown Prosecution Service, Consents to Prosecute. 31 October 2018, updated 11 December 2018. Available at: https://www.cps.gov.uk/legal-guidance/consents-prosecute (last visited 13 July 2021).
