Saturday, 11 March 2023

Five lessons from the Loi Avia

In a few months’ time three years will have passed since the French Constitutional Council struck down the core provisions of the Loi Avia - France’s equivalent of the German NetzDG law – for incompatibility with fundamental rights. Although the controversy over the Loi Avia has passed into internet history, the Constitutional Council's decision provides some instructive comparisons when we examine the UK’s Online Safety Bill.

As the Bill awaits its House of Lords Committee debates, this is an opportune moment to cast our minds back to the Loi Avia decision and see what lessons it may hold. Caution is necessary in extrapolating from judgments on fundamental rights, since they are highly fact-specific; and when they do lay down principles they tend to leave cavernous room for future interpretation. Nevertheless, the Loi Avia decision makes uncomfortable reading for some core aspects of the Online Safety Bill.


The key features of the Loi Avia were

  • For illegal CSEA and terrorism content, one hour removal of content notified to an in-scope publisher or host by the administrative authority, on pain of one year’s imprisonment and a 250,000 euro fine.

The Constitutional Council’s objection was founded on the determination of illegality being at the sole discretion of the administrative authority. This provision has no direct parallel in the Online Safety Bill. However, similar considerations could come into play should an Ofcom Code of Practice recommend giving state agencies some kind of trusted flagger status.

  • For content contravening specified hate-related, genocide-related, sexual harassment and child pornography laws, 24-hour removal of manifestly illegal content following notification by any person to an in-scope platform operator, under penalty of a fine of 250,000 euros.

The Online Safety Bill analogue is a reactive ‘swift take down’ duty on becoming aware of in-scope illegal content. Unlike the Loi Avia, the Bill also imposes proactive prevention duties.

The Online Safety Bill imposes duties for both illegal content and legal content harmful to children. Since the Loi Avia concerned only illegal content, the Constitutional Council did not have to consider obligations relating to ‘legal but harmful’ content of any kind, whether for adults or children.

Lesson 1: The rule of law comes first

The tests that the Constitutional Council applied to the Loi Avia – legality, necessity and proportionality – are components of the European Convention on Human Rights, with which the Online Safety Bill must comply.

Along the obstacle course of human rights compatibility, the first hurdle is legality: known in the ECHR as the “prescribed by law” test. In short, a law must have the quality of law to qualify as law. If the law does not enable someone to foresee with reasonable certainty whether their proposed conduct is liable to be affected as a consequence of the law, it falls at that first hurdle. If legislation will result in arbitrary or capricious decisions - for example through vagueness or grant of excessive discretion - it lacks the essential quality of law.

The problem with vagueness was spelt out by the House of Lords in R v Rimmington, citing the US case of Grayned:

"Vagueness offends several important values … A vague law impermissibly delegates basic policy matters to policemen, judges and juries for resolution on an ad hoc and subjective basis, with the attendant dangers of arbitrary and discriminatory application."

Whilst most often applied to criminal liability, the legality objection has also been described as a constitutional principle that underpins the rule of law generally. Lord Diplock referred to it in a 1975 civil case (Black-Clawson):

"The acceptance of the rule of law as a constitutional principle requires that a citizen, before committing himself to any course of action, should be able to know in advance what are the legal consequences that will flow from it."

The French Constitutional Council held that the Loi Avia failed the legality test in one respect. The Loi provided that the intentional element of the offence of failure to remove content notified by any person could arise from absence of a “proportionate and necessary examination of the notified content”. The Constitutional Council found that if this was intended to provide a defence for platform operators, it was not drafted in terms that allowed its scope to be determined. In other words, a defence (if that is what it was) of having carried out a proportionate and necessary examination was too vague to pass the legality test.

The Online Safety Bill differs from the Loi Avia. It does not impose criminal liability on a platform for failure to take down a particular item of user content.  Enforcement by the appointed regulator, Ofcom, is aimed at systematic failures to fulfil duties rather than at individual content decisions. Nevertheless, the Bill is liberally sprinkled with references to proportionality – similar language to that which the French Constitutional Council held was too vague. It typically couches platform and search engine duties as an obligation to use proportionate systems and processes designed to achieve a stipulated result.

It is open to question whether compliance with the legality principle can be achieved simply by inserting ‘proportionate’ into a broadly stated legal duty, instead of grasping the nettle of articulating a more concrete obligation that would enable the proportionality of the interference with fundamental rights to be assessed by a court.

The government’s ECHR Memorandum seeks to head off any objection along these lines by stressing the higher degree of certainty that it expects would be achieved when Ofcom’s Codes of Practice have been laid before Parliament and come into effect. Even if that does the trick, it is another matter whether it is desirable to grant that amount of discretion over individual speech to a regulator such as Ofcom.

For the Online Safety Bill the main relevance of the legality hurdle is to the freedom of expression rights of individual users. Can a user foresee with reasonable certainly whether their proposed communication is liable to be affected as a result of a platform or search engine seeking to fulfil a safety duty imposed by the legislation? The Bill requires those online intermediaries to play detective, judge and bailiff. Interpolation of an online intermediary into the process of adjudging and sanctioning user content is capable of introducing arbitrariness that is not present when the same offence is prosecuted through the courts, with their attendant due process protections.

In the case of the Online Safety Bill, arbitrariness is a real prospect. That is largely because of the kinds of offences on which platforms and search engines are required to adjudicate, the limited information available to them, and the standard to which they have to be satisfied that the user content is illegal.

Lesson 2: Beyond ‘manifestly illegal’

An intriguing feature of the Constitutional Council decision is that although the Loi Avia prescribed, on the face of it, a high threshold for removal of illegal content – manifest illegality  that was not enough to save the legislation from unconstitutionality.  ‘Manifestly illegal’ is a more stringent test than the ‘reasonable grounds to infer’ threshold prescribed by the Online Safety Bill.

The Loi Avia required removal of manifestly illegal user content within 24 hours of receiving from anyone a notification which gave the notifier’s identity, the location of the content, and which specified the legal grounds on which the content was said to be manifestly illegal.

The Constitutional Council observed that the legislation required the operator to examine all content reported to it, however numerous the reports, so as not to risk being penalised. Moreover, once reported the platform had to consider not only the specific grounds on which the content was reported, but all offences within the scope of the legislation – even though some might present legal technicalities or call for an assessment of context. These issues were especially significant in the light of the 24 hour removal deadline and the criminal penalty for each failure to withdraw.

In the Constitutional Council’s view the consequence of these provisions, taking into account also the absence of any clearly specified defence to liability, was that operators could only be encouraged to withdraw content reported to them, whether or not it was manifestly illegal. That was not necessary, appropriate or proportionate and so was unconstitutional.

The Online Safety Bill does not prescribe specific time limits, but requires swift removal of user content upon the platform becoming aware of in-scope illegality. As with the Loi Avia, that applies to all in-scope offences.

The touchstone for assessment of illegality under the Bill is reasonable grounds to infer illegality, on the basis of all information reasonably available to the platform. Unless that threshold is surmounted, the platform does not have to remove it. If it is surmounted, the platform must do so swiftly.

At least in the case of automated proactive monitoring and filtering, the available information will be minimal – the users’ posts themselves and whatever the system knows about the relevant users. As a consequence, the decisions required to be made for many kinds of offence – especially those dependent on context - will inevitably be arbitrary. Moreover, a platform has to ignore the possibility of a defence unless it has something from which it can infer on reasonable grounds that a defence may succeed.

Whilst the Online Safety Bill lacks the Loi Avia’s chilling sword of Damocles of short prescriptive deadlines and automatic criminal liability for failure to remove, the reason why those factors (among others) were legally significant was their effect on the freedom of expression of users: the likely over-removal of lawful user content. The Online Safety Bill’s lower threshold for adjudging illegality, combined with the requirement to make those judgments in a relative information vacuum - often at scale and speed - does more than just encourage takedown of legal user content: it requires it.

Lesson 3 –The lens of prior restraint

The briefly glimpsed elephant in the room of the Loi Avia decision is prior restraint. The Constitutional Council alluded to it when it remarked that the removal obligations were not subject to the prior intervention of a judge or subject to any other condition.

Legislation requiring a platform summarily to adjudge the legality of individual items of user content at speed and at scale bears the hallmarks of prior restraint: removal prior to full adjudication on the merits after argument and evidence.

Prior restraint is not impermissible. It does require the most stringent scrutiny and circumscription, in which the risk of removal of legal content will loom large. The ECtHR in Yildirim considered an interim court order blocking Google Sites.  It characterised that as a prior restraint, and observed: “the dangers inherent in prior restraints are such that they call for the most careful scrutiny on the part of the Court”. 

The ECtHR in Animal Defenders v UK distinguished a prior restraint imposed on an individual act of expression from general measures: in that case a ban on broadcasting political advertising.

If an individual item is removed ultimately pursuant to a general measure, that does not prevent the action being characterised as a prior restraint. If it did, the doctrine could not be applied to courts issuing interim injunctions. The fact that the Online Safety Bill does not penalise a platform for getting an individual decision wrong does not disguise the fact that the required task is to make judgments about individual items of user content constituting individual acts of expression.

The appropriateness of categorising at least proactive detection and filtering obligations as a form of prior restraint is reinforced by the CJEU decision in Poland v The European Parliament and Council, which applied Yildirim to those kinds of provisions in the context of copyright.

Lesson 4 – Context, context, context

The Constitutional Council pointed out the need to assess context for some offences. That is all the more significant for the Online Safety Bill, for several reasons.

First, unlike the Loi Avia the Online Safety Bill imposes proactive, not just reactive, duties. That will multiply the volume of user content to be assessed, in many cases requiring the deployment of automated content monitoring. Such systems, by their very nature, can be aware only of content flowing through the system and not of any external context. 

Second, the Bill requires illegality assessments to be made ignoring external contextual information unless it is reasonably available to the platform.

Third, defences such as reasonableness will often be inherently contextual. The Bill, however, enables the intermediary to take account of the possibility of a defence only if it has information on the basis of which it can infer that a defence may successfully be relied upon.  

Lesson 5 – Proactive duties

The Loi Avia decision was about reactive duties based on notification. Proactive illegality duties present inherently greater human rights challenges. A less prescriptive, less draconian reactive regime, combined with a ‘manifest illegality’ standard and greater due process safeguards, might possibly have survived. But if the starting point is aversion to a regime that encourages takedown of legal user content, it is difficult to see how a regime that carries a certainty of over-takedown, as do the Online Safety Bill’s proactive illegality duties, could pass muster.

What is to be done?

Raising the Online Safety Bill’s standard of assessment from reasonable grounds to infer to manifest illegality would go some way towards a better prospect of human rights compliance. But that still leaves the problem of the assessment having to be made in ignorance of external context; and the problem of the possibility of a defence being discounted unless it is apparent from the information flowing through the system. Those more intractable issues put in question the kinds of offences that platforms and search engines could be called upon to adjudge. 

Tuesday, 24 January 2023

Positive light or fog in the Channel?

If anything graphically illustrates the perilous waters into which we venture when we require online intermediaries to pass judgment on the legality of user-generated content, it is the government’s decision to add S.24 of the Immigration Act 1971 to the Online Safety Bill’s list of “priority illegal content”: user content that platforms must detect and remove proactively, not just by reacting to notifications. Proactive measures could involve scouring the platform for content already uploaded, filtering and blocking at the point of attempted upload, or both.

The political target of the Bill amendment, which the government says it will introduce in the House of Lords, is videos of migrants crossing the Channel in boats. The Secretary of State explained it thus:
“We will also add Section 24 of the Immigration Act 1971 to the priority offences list in Schedule 7. Although the offences in Section 24 cannot be carried out online, paragraph 33 of the Schedule states that priority illegal content includes the inchoate offences relating to the offences that are listed. Therefore aiding, abetting, counselling, conspiring etc those offences by posting videos of people crossing the channel which show that activity in a positive light could be an offence that is committed online and therefore falls within what is priority illegal content. The result of this amendment would therefore be that platforms would have to proactively remove that content.”
We have to assume that this wheeze was dreamed up in some haste, meeting the immediate political imperative to respond to a strongly supported back bench amendment that tried to tack videos of boat crossings on to the Bill’s children’s duties. Now that the dust has settled, at least temporarily, let us take a look at what would be involved in applying the government's proposal.

In view of some of the media commentary, it is worth emphasising that the proposed amendment to the Bill would not create a new offence. It is based on existing accessory liability legislation, which platforms (and indeed search engines) would have to apply proactively.

In a positive light

Where does ‘in a positive light’ come from? Presumably the Secretary of State must have had in mind that if a video shows the activity of crossing the Channel to gain illegal entry to the UK in a negative light – thus tending to deter the activity - that cannot amount to counselling (in modern language, encouraging) an offence of entering (or attempting to enter) the UK illegally. So far so good. But that does not mean we should jump to the conclusion that ‘in a positive light’ is sufficient to amount to encouragement.

The offence of aiding, abetting, counselling etc a Section 24 offence applies not only to videos but to any kind of communication, whether on social media, simple discussion forums, websites or elsewhere.

You do not have to go far to find studies suggesting that illegal immigration can have positive benefits to an economy. Does supporting that position in an online discussion about UK immigration put the activity of illegal entry to the UK in a positive light? Quite possibly. Does it (in the legal sense) encourage an offence of illegal entry to the UK? Surely not. That is a far cry from intentionally encouraging a prospective illegal migrant to commit an illegal entry offence.

The idea that someone might be prosecuted for voicing that kind of opinion in a general online discussion is (one would hope) absurd. It brings to mind the comment of Lord Scott in Rusbridger v Attorney-General, a case about the moribund Section 3 of the Treason Felony Act 1848:
“[Y]ou do not have to be a very good lawyer to know that to advocate the abolition of the monarchy and its replacement by a republic by peaceful and constitutional means will lead neither to prosecution nor to conviction. All you need to be is a lawyer with commonsense.”
In any event legislation must, so far as it is possible to do so, be read and given effect in a way which is compatible with the European Convention on Human Rights right of freedom of expression (S.3 Human Rights Act 1998; albeit the Bill of Rights Bill would repeal that provision).

The Secretary of State’s proposal has reportedly sparked fears among humanitarian organisations of consequences if they share footage that may call into question the policing of Channel crossings. The Home Office, for its part, has said that they would not be penalised. That is an understandable view if all the legal elements of an encouragement offence are properly taken into account.

Nevertheless, it is not so far-fetched a notion that an online platform, tasked by the Online Safety Bill proactively to detect and remove user content that encourages an illegal entry offence, might consider itself duty-bound to remove content that in actual fact would not result in prosecution or a conviction in court. There are specific reasons for this under the Bill, which contrast with prosecution through the courts.

Prosecution versus the Bill's illegality duties

First the platform’s removal duty under the Bill kicks in not if the user’s content is illegal beyond reasonable doubt, or manifestly illegal, but if the platform has ‘reasonable grounds to infer’ illegality – on the face of it a significantly lower standard. Whether this standard is compatible with Article 10 of the European Convention on Human Rights is questionable, but nevertheless it is what the Bill says. The Bill would inevitably require platforms to remove some content that is in fact legal.

Second, the Bill requires platforms to act on all the information reasonably available to the platform: a far more limited factual basis than a court. At least for an automated system that would be likely to be the content of the post and any related information on the platform (such as information indicating the nature and identity of the poster). It excludes any extrinsic contextual information not reasonably available to the platform. 

Further, the platform can take into account the possibility of a defence only if it has reasonable grounds to infer that one may successfully be relied upon. For many defences (such as reasonable excuse) any grounds for a defence will not necessarily be apparent from the information available to the platform, in which case the possibility of a defence must be ignored. 

The platform’s assessment of illegality may thus depend on the happenstance of whether there is anything in the post itself, or its surrounding data, that points to the possibility of a successful defence. For some widely drawn offences intent and available defences are the most significant elements in determining legality, and are integral to the balance drawn by the legislature. This, we shall see, is of particular relevance to the encouragement and assistance offences under the Serious Crime Act 2007.

Third, the task of a platform is not to second-guess whether the authorities would prosecute, but to decide whether it has reasonable grounds to infer that the content falls within the letter of the law. Whilst the Bill makes numerous references to proportionality, that does not affect the basis on which the platform must determine illegality. That is a binary, yes or no assessment. There is no obvious room for a platform to conclude that something is only a little bit illegal, or to decide that, once detected, some content crossing the ‘reasonable grounds to infer’ threshold could be left up. Certainly the political expectation is that any detected illegal content will be removed.

If that is right, the assessment that platforms are required to make under the Bill lacks the anything akin to the ameliorating effect of prosecutorial discretion on the rough edges of the criminal law. Conversely to build such discretion, even principles-based, into the decision-making required of platforms would hardly be a solution either, especially not at the scale and speed implied by automated proactive detection and removal obligations. We do not want platforms to be arbiters of truth, but to ask them (or their automated systems) to be judges of the public interest or of the seriousness of offending would be a recipe for guesswork and arbitrariness, even under the guidance of Ofcom.

If this seems like a double bind, it is. It reflects a fundamental flaw in the Bill’s duty of care approach: the criminal law was designed to be operated within the context of the procedural protections provided by the legal system, and to be adjudged by courts on established facts after due deliberation; not to be the subject of summary justice dispensed on the basis of incomplete information by platforms and their automated systems tasked with undertaking proactive detection.

Fourth, we shall see that in some cases the task required of the platform appears to involve projection into the future on hypothetical facts. Courts are loath to assess future criminal illegality on a hypothetical basis. Their task at trial is to determine whether the events that are proved in fact to have occurred amounted to an offence.

Fifth, inaccuracy. False positives are inevitable with any moderation system - all the more so if automated filtering systems are deployed and are required to act on incomplete information (albeit Ofcom is constrained to some extent by considerations of accuracy, effectiveness and lack of bias in its ability to recommend proactive technology in its Codes of Practice). Moreover, since the dividing line drawn by the Bill is not actual illegality but reasonable grounds to infer illegality, the Bill necessarily deems some false positives to be true positives.

Sixth, the involvement of Ofcom. The platform would have the assistance of a Code of Practice issued by Ofcom. That would no doubt include a section describing the law on encouragement and assistance in the context of the S.24 1971 Act illegal entry offences, and would attempt to draw some lines to guide the platform’s decisions about whether it had reasonable grounds to infer illegality.

An Ofcom Code of Practice would carry substantial legal and practical weight. That is because the Bill provides that taking the measures recommended in a Code of Practice is deemed to fulfil the platform’s duties under the Bill. Much would therefore rest on Ofcom’s view of the law of encouragement and assistance and what would constitute reasonable grounds to draw an inference of illegality in various factual scenarios.

Seventh, the involvement of the Secretary of State. Ofcom might consider whether to adopt the Secretary of State’s ‘in a positive light’ interpretation. As the Bill currently stands, if the Secretary of State did not approve of Ofcom’s recommendation for public policy reasons s/he could send the draft Code of Practice back to Ofcom to with a direction to modify – and, it seems, keep on doing so until s/he was happy with its contents.

Even if that controversial power of direction were removed from the Bill, Ofcom would still have significant day to day power to adopt interpretations of the law and apply them to platforms’ decision-making (albeit Ofcom’s interpretations would in principle be open to challenge by judicial review).

As against those seven points, in fulfilling its duties under the Bill a platform is required to have particular regard to the importance of protecting users’ right to freedom of expression within the law. ‘Within the law’ might suggest that the duty has minimal relevance to the illegality duties, especially when clause 170 sets out expressly how platforms are to determine illegality. It provides that if the reasonable grounds to infer test is satisfied, the platform must treat the content as illegal.

The government’s ECHR Memorandum suggests that the ‘have particular regard’ duty may have some effect on illegality determination, but it does not explain how it does so in the face of the express provisions of clause 170. It also inaccurately paraphrases clause 18 by omitting ‘within the law’:
“34. Under clause 18, all in-scope service providers are required to have regard to the importance of protecting freedom of expression when deciding on and implementing their safety policies and procedures. This will include assessments as to whether content is illegal or of a certain type and how to fulfil its duties in relation to such content. Clause 170 makes clear that providers are not required to treat content as illegal content (i.e. to remove it from their service) unless they have reasonable grounds to infer that all elements of a relevant offence are made out. They must make that inference on the basis of all relevant information reasonably available to them.”
That is all by way of lengthy preliminary. Now let us delve into how a platform might be required to go about assessing the legality of a Channel dinghy video under the Accessories and Abettors Act 1861, then for the companion encouragement and assistance offences under the Serious Crime Act 2007.

Let us assume that the Secretary of State is right: that posting a video of people crossing the Channel in dinghies, which shows that activity in a positive light, can in principle amount to encouraging an illegal entry offence. In the interests of simplicity, I will ignore the Secretary of State’s reference to conspiracy. How should a platform go about determining illegality?

Spoiler alert: the process is more complicated and difficult than the Secretary of State’s pronouncement might suggest. And in case anyone is inclined to charge me with excessive legal pedantry, let us not forget that the task that the Bill expressly requires a platform to undertake is to apply the rules laid down in the Bill and in the relevant underlying offences. The task is not to take a rough and ready ‘that looks a bit dodgy, take it down’, or ‘the Home Secretary has complained about this content so we’d better remove it’ approach. Whether what the Bill requires is at all realistic is another matter.

Aiding, abetting and counselling – the 1861 Act

Aiding, abetting and counselling (the words used by the Secretary of State) is the language of the 1861 Act: “Whosever shall aid, abet, counsel or procure the commission of any indictable offence … shall be liable to be tried, indicted and punished as a principal offender.”

One of the most significant features of accessory liability under the 1861 Act is that there can be no liability for aiding, abetting, counselling or procuring unless and until the principal offence has actually occurred. Whilst the aiding, abetting etc does not have to cause the principal offence that occurred, there has to be some connecting link with it. As Toulson LJ put it in Stringer:
“Whereas the provision of assistance need not involve communication between D and P, encouragement by its nature involves some form of transmission of the encouragement by words or conduct, whether directly or via an intermediary. An un-posted letter of encouragement would not be encouragement unless P chanced to discover it and read it. Similarly, it would be unreal to regard P as acting with the assistance or encouragement of D if the only encouragement took the form of words spoken by D out of P's earshot.”
Timing This gives rise to a timing problem for a platform tasked with assessing whether a video is illegal. For illegality to arise under the 1861 Act the video must in fact have been viewed by someone contemplating an illegal entry offence, the video would have to have encouraged them to enter the UK illegally, and they would have to have proceeded to do so (or attempt to do so).

Absent those factual events having taken place, there can be no offence of aiding and abetting. The aiding and abetting offence would further require the person posting the video to have intended the person contemplating illegal entry to view the video and to have intended to encourage their actual subsequent actual or attempted illegal entry.

Thus if a platform is assessing a video that is present on the platform, in order to adjudge the video to be illegal it would at a minimum have to consider how long it has been present on the platform. That is because there must be reasonable grounds to infer both that a prospective migrant has viewed it and that since doing so that person has already either entered the UK illegally or attempted to do so. Otherwise no principal offence has yet occurred and so no offence of aiding and abetting the principal offence can have been committed by posting the video.

It may in any case be a nice question whether, in the absence of any evidence available to the platform that a prospective migrant has in fact viewed the video, the platform would have reasonable grounds to infer the existence of any of these facts. To do so would appear to involve making an assumption of someone viewing the video and of a connected illegal entry offence that the assumed viewing has in fact encouraged. 

For a post blocked by filtering at the point of upload (if that were considered feasible) the timing issue becomes a conundrum. Since no-one can have viewed a blocked video, none of the required subsequent events can possibly have occurred. Nor does the law provide any offence of attempting to aid and abet a 1971 Act offence.

Thus at least for upload filtering it appears that either there is a conceptual bar to a platform determining that a video blocked at the point of upload amounts to aiding abetting, or the platform would (if the Bill permits it) have to engage in some legal time travel and assess illegality on a hypothetical future basis.

A basis on which a platform could be required to assess such hypothetical illegality may be provided by Clause 53(14)(b) of the Bill, which in effect provides that illegal content includes content that would be illegal if it were present on the platform. 

Even then, a video present on the platform only as a legal fiction cannot as a matter of fact be connected to any subsequent actual encouraged primary offence. Deemed presence would therefore have to be notionally extended for a sufficient period to hypothesise the factual events necessary for completion of the aiding and abetting offence: that a notional prospective migrant has hypothetically viewed the video present on the service, hypothetically been encouraged by the video to commit or attempt an illegal entry offence, and hypothetically then done so.

Even if any of this hypothesising is permissible under the Bill, whether it could provide reasonable grounds to infer illegality is a matter for conjecture. The need to hypothesise the existence of an actual illegal entry offence would never arise in a prosecution in court, since for a prosecution of the accessory to succeed it must be proved that the principal offence has taken place. In court, therefore, the assessment of accessory liability will always be within the context of a known past set of facts that are proved to have amounted to an offence by a principal.

Intent The platform would also have to consider whether it has reasonable grounds to infer that the poster had the necessary intention to aid, abet etc the actual or attempted offence.

In court the prosecution would have to prove, beyond reasonable doubt, that the poster intended a viewer of the video to obtain or attempt illegal entry to the UK, the poster having knowledge of the facts that would and did render the principal’s conduct criminal. (‘Did’, because there can be no conviction for aiding and abetting unless the principal offence is proved to have taken place.)

That would raise the question of whether generalised knowledge of the existence of people crossing the Channel who might view the video and be encouraged by it would be sufficient to satisfy the knowledge requirement, when the poster would have been unaware of the particular individual who had in fact viewed the video and then committed the offence. Whilst it might be legitimate to find intent where the video is specifically promoting illegal crossings to prospective migrants, such a finding would seem to be highly debatable if the video did not offer targeted encouragement, even if it portrayed such activities in a positive light.

How should a platform decide whether the poster of the video had the requisite intent to constitute an aiding and abetting offence? The Bill requires the platform to apply the ‘reasonable grounds to infer’ test. It has to make that assessment on the basis of all the information reasonably available to it. That would likely bring in to account not only the content of the video, but any surrounding text in the post and (if apparent) the nature of the person posting. The intent of a video advertising illegal Channel crossings might be clear, the intent of a bare clip of a dinghy carrying migrants (even if it showed smiling occupants and was accompanied by upbeat music) not so much.

Serious Crime Act 2007 – encouraging and assisting

We started by considering aiding and abetting under the 1861 Act because that is what the language used by the Secretary of State appeared to allude to. That is not, however, the end of the story. The Serious Crime Act 2007 enacted encouragement and assistance offences that, unlike aiding and abetting, do not depend on the principal offence actually taking place. They therefore avoid the time travel and hypothesising contortions involved in applying the Bill to the 1861 Act.

Also unlike aiding and abetting, an attempt to commit an encouragement or assistance offence under the 2007 Act is itself an offence. In principle therefore, a foiled attempt to upload a video capable of constituting an encouragement or assistance offence under the 2007 Act could itself constitute an offence.

By way of illustration, consider the simplest 2007 Act offence, S.44:
“(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.”
So a platform tasked with adjudging whether the video is illegal would have to consider not only whether posting the video is ‘capable’ of encouraging the commission of an unlawful entry offence, but also whether the person who posted it intended to encourage the commission of the offence; bearing in mind that a mere foreseeable consequence does not count as intent. (That, it might be thought, rules out any but the most targeted advertising or promotional videos.)

How should a platform go about these two tasks? As with the 1861 Act aiding and abetting offences, part of the answer lies in Clause 170 of the Bill, which specifies the standard of ‘reasonable grounds to infer’ based on ‘all information reasonably available’ to the platform.

The analysis would be based on the same information as for aiding and abetting, but without the need to show (or hypothesise) that anyone actually viewed or acted upon the video. It is enough if publication of the video is capable of encouraging the offence. However, the express exclusion of a merely foreseeable consequence would limit the inference of intention that it is reasonable for the platform to draw.

Defence of reasonable conduct Unlike for the 1861 Act aiding and abetting offence, the 2007 Act offences provide a defence of ‘reasonable conduct’. This comes in two different versions:

(1) that the defendant knew that certain circumstances existed and that it was reasonable for him to act as he did in those circumstances; or

(2) that he believed certain circumstances to exist, that his belief was reasonable, and that it was reasonable for him to act as he did in the circumstances as he believed them to be.

Factors that the 2007 Act states have to be considered in relation to reasonableness include the seriousness of the offence and any purpose for which the defendant claims to have been acting. A 2007 Act defence will succeed in court if the defendant proves it on the balance of probabilities.

The information on which the possibility of a reasonableness defence depends may well be extrinsic to the platform or its automated systems. The purpose for which a user has acted is something within the user’s knowledge and belief and may not be apparent from the post itself.

As already mentioned, this is significant because the platform cannot consider the possibility of a defence unless, on the basis of all relevant information that is reasonably available to it, it has reasonable grounds to infer that a defence may be successfully relied upon (in the context of the 2007 Act defence: successful on the balance of probabilities).

In determining what information is reasonably available to the provider, the following factors, in particular, are relevant: (a) the size and capacity of the provider, and (b) whether a judgement is made by human moderators, by means of automated systems or processes or by means of automated systems or processes together with human moderators.

The probable net result, for an automated system, is that the possibility of a defence is to be ignored unless it is apparent from the information processed by the system. Yet for the 2007 Act encouragement and assistance offences, the defences are an integral element of the offence, designed to balance the potentially overreaching effects of inchoate liability founded on mere capability.

In reality, however, it smacks of fantasy to imagine that a platform, whether employing automated systems, human moderators, or a combination of the two, would be capable of applying rules of this nuance and complexity, particularly in real or near real time.

The broader issue

These problems with the Bill’s illegality duties are not restricted to migrant boat videos or immigration offences, although the Secretary of State’s statement has provided an unexpected opportunity to illustrate them. They are of general application and are symptomatic of a flawed assumption at the heart of the Bill: that it is a simple matter to ascertain illegality just by looking at what the user has posted. There will be some offences for which this is possible (child abuse images being the most obvious), and other instances where the intent of the poster is clear. But for the most part that will not be the case, and the task required of platforms will inevitably descend into guesswork and arbitrariness: to the detriment of users and their right of freedom of expression.

It is strongly arguable that if an illegality duty is to be placed on platforms at all, the threshold for illegality assessment should not be ‘reasonable grounds to infer’, but clearly or manifestly illegal. Indeed, that may be what compatibility with the Article 10 right of freedom of expression requires.

Friday, 6 January 2023

Twenty questions about the Online Safety Bill

Before Christmas Culture Secretary Michelle Donelan invited members of the public to submit questions about the Online Safety Bill, which she will sit down to answer in the New Year. 

Here are mine. 

1. A volunteer who sets up and operates a Mastodon instance in their spare time appears to be the provider of a user-to-user service. Is that correct?

2. Alice runs a personal blog on a blogging platform and is able to decide which third party comments on her blogposts to accept or reject. Is Alice (subject to any Schedule 1 exemptions) the provider of a user-to-user service in relation to those third party comments?

3. Bob runs a blog on a blogging platform. He has multiple contributors, whom he selects. Is Bob the provider of a user-to-user service in relation to their contributions?

4. Is a collaborative software development platform the provider of a user-to-user service?

5. The exclusion from “regulated user-generated content” extends to comments on comments (Clause 49(6)). But a facility enabling free form ‘comments on comments’ appears to disapply the Sch 1 para 4 limited functionality user-to-user service exemption. Is that correct? If so, what is the rationale for the difference? Would, for example, a newspaper website with functionality that enabled free form ‘comments on comments’ therefore not enjoy exclusion from scope under Sch 1 para 4?

6. Does the Sch 1 para 4 limited functionality exemption apply to goods retailers’ own-product review sections? If so, does it achieve that when it refers only to content and not to the goods themselves?

7. Would a site that enables academics to upload papers, subject to prior review by the site operator, be a user-to-user service? 

8. Cl 204(2)(e) appears to suggest that a multiplayer online game would be a user-to-user service by virtue of player interaction alone, whether or not there is an inter-player chat or similar facility. Is that right?

9. Carol sets up and operates a voluntary online neighbourhood watch forum for her locality. Would Carol be a provider of a user-to-user service? 

10. Dan operates a blockchain node. Would Dan be a provider of a user-to-user service?

11. Grace chairs a public meeting using a video platform. Grace has control over who can join the meeting. Would Grace be a provider of a user-to-user service in relation to that meeting?

12. The threshold that the Bill requires a platform to apply when determining criminal illegality is ‘reasonable grounds to infer’. The criminal standard of proof is ‘beyond reasonable doubt’. Would not the Bill’s lower threshold inevitably require removal (at least for proactive obligations) of content that is in fact legal? For automated real time systems would that not occur at scale?

13. The Bill requires a platform to adjudge illegality on the basis of all relevant information reasonably available to it. Particularly for proactive automated processes, that will be limited to what users have posted to the platform. Yet often, illegality depends crucially on extrinsic contextual information that is not available to the platform. How could the adjudgment required by the Bill thus not be arbitrary?

14. For many offences the question of illegality is likely to revolve mainly around intent and available defences. The Bill requires platforms to assess illegality on the basis that the possibility of a defence is to be taken into account only if the platform has reasonable grounds to infer that a defence may successfully be relied upon. Yet the information from which the possibility of a defence (such as reasonable excuse) might be inferred will very often be extrinsic context that, especially for proactive obligations, is not available to a platform. Would that not inevitably require removal of content that is in fact legal? For automated real time systems would that not occur at scale?

15. The Bill requires platforms to have particular regard to the importance of protecting users’ right to freedom of expression ‘within the law’. Does that modify the express requirements of Clause 170 as to how a platform should assess illegality? If so, how?

16. The government’s European Convention on Human Rights Memorandum contains no discussion of the Bill’s illegality duties as a form of prior restraint. Nor does it address the human rights implications of the ‘reasonable grounds to infer’ clause, which was introduced later. Will the government issue a revised Memorandum?

17. Is it intended that the risks of harm to individuals to be mitigated and managed under Clause 9(2)(c) should be limited to those arising from illegality identified in the illegality risk assessment? If so, how does the Bill achieve that?

18. The Bill contains powers to require private messaging services to use accredited technology to identify CSEA content. It also contains an obligation to report all new detected material to the National Crime Agency. The Explanatory Notes state that services will be required to report all and any available information relating to instances of CSEA, including any that help identify a perpetrator or victim. 

The White Paper noted that “Many children and young people take and share sexual images. Creating, possessing, copying or distributing sexual or indecent images of children and young people under the age of 18 is illegal, including those taken and shared by the subject of the image.” Does this mean that an under-18 consensually taking and sharing an indecent selfie on a private messaging platform would automatically be reported to the National Crime Agency if the image is detected by the platform?

19. What are the estimated familiarisation and compliance costs for an in-scope small business or voluntary user-to-user service? What is the calculation of the estimated costs? 

20. The Law Commission in 2018 stated that the common law public nuisance offence applied to online communications. The statutory replacement in s.78 of the Police, Crime, Sentencing and Courts Act 2022 does so too. Could a platform’s reactive duty under Cl. 9, combined with Cl. 170, require it to determine whether it has reasonable grounds to infer that a user’s post creates a risk of causing serious annoyance to a section of the public?