Safety by design or systems for content moderation?

The Online Safety Act Network (OSAN) recently published a 10-point plan to amend the Online Safety Act. The plan includes:

“Insert a definition of safety by design into the Act to make clear to Ofcom and services what Parliament intended”.

From a technical drafting perspective clarification might be welcome. The Act says that it “seeks to secure that regulated services are safe by design”. That was added at the last minute in a clause describing the overall purpose of the legislation, but which lacked any definition of ‘safe’ or ‘safe by design’. I discussed here the undoubted difficulties in interpreting what is now Section 1 of the Act.

Of course, before we can craft a definition of safety by design we have to know what it is intended to mean. For myself, I have always regarded much of the theory underlying safety by design as fragile, at least within the context of the Online Safety Act. But putting those doubts on one side, I did think that I had a reasonable idea of what safety by design was intended to be about.

Now I am not so sure.

To recap, this is how I thought safety by design was meant to apply to regulation of online platforms:

• Safety by design requires safety to be considered at the design stage, not as an afterthought.

• Safety by design should then be applied iteratively via periodic risk assessments, incorporating feedback learned during operation of the service.

• Safety by design focuses on platform systems and processes, identifying and addressing those that create or exacerbate a risk of harm (however defined).

• Safety by design is not, or at least not primarily, about systems for content moderation.

• Safety by design favours non-content-specific, systems-focused measures.

• Safety by design is not about automated content detection and filtering.

Safety by design proponents have long criticised the Online Safety Act for being too content-focused. Rather than more and better content moderation, platforms should have to design safety into their systems and processes from the outset. This, so the theory goes, would result in less harm (however that might be conceptualised) occurring on platforms and less need for ex-post content moderation.

There have been variations on these themes: for instance, that a systems focus can include friction measures targeted at specific kinds of content, but which stop short of requiring removal; or that measures can focus on harm arising from certain kinds of content without focusing specifically on the content itself.

Nevertheless, as a general proposition I understood safety by design (a.k.a. ‘systems and processes’) to be about addressing from the outset the design of risk-creating system features, combined with a preference for non-content-related or content-agnostic measures over content-specific measures.

If that is right, safety by design has two elements. It articulates a general approach to safety but is also exclusionary: systems for content moderation (or at least automated filtering systems) are not a safety by design measure. The UK government, it should be said, has taken the opposite view. It regards automated content filtering as a safety by design measure. That is the most obvious difference of view that, if I am right in my understanding of safety by design, would have to be resolved in crafting a statutory definition.

To the extent that safety by design proponents embrace systems for content moderation, that has tended to be as a fall-back for where safety by design measures have not squeezed harm out of the system.

Thus Professor Lorna Woods’ October 2024 paper for OSAN, Safety by Design, although allowing for the possibility of ex post measures as a residual measure, differentiated that from a primary focus on design choices:

“At the moment, content moderation seems to be in tension with the design features that are influencing the creation of the content in the first place, making moderation a harder job. So a ‘by design’ approach is a necessary precondition for ensuring that other ex post responses have a chance of success.

While a “by design” approach is important, it is not sufficient on its own; there will be a need to keep reviewing design choices and updating them, as well as perhaps considering ex post measures to deal with residual issues that cannot be designed out, even if the incidence of such issues has been reduced.”

She distinguished safety by design from techno-solutionism:

“Designing for safety (or some other societal value) does not equate to techno-solutionism (or techno-optimism); the reliance on a “magic box” to solve society’s woes or provide a quick fix. Rather, what it acknowledges is that each technology may have weaknesses and disadvantages, as well as benefits. Further, the design may embody the social values and interests of its creators. A product (or some of its features) may be part of the problem. The objective of “safety by design” is – like product safety – to reduce the tendency of a given feature or service to create or exacerbate such issues.” (emphasis added)

One might think that automated content filtering is the paradigm example of regulatory techno-solutionism. Indeed, as to the Online Safety Act itself, Professor Woods noted its emphasis on systems for content moderation:

“What is rather more explicit in the [Online Safety Act] safety duties is the focus on filtering and moderation, which may have a design element (i.e. the tools are made available within the system and designed to work with the system) but seem more ex post in the way they work.”

Elsewhere Professor Woods has included reactive content take-down systems within safety by design, but as the “last port of call”. (Introducing the Systems Approach and the Statutory Duty of Care (chapter in Perspectives on Platform Regulation, Nomos, 2021).)

We can find other examples of safety by design proponents expressing concern about the Online Safety Act’s focus on systems for content moderation.

Carnegie UK was OSAN’s online safety policy predecessor and, through the work of Professor Woods and William Perrin, was the originator of the proposal for a statutory duty of care. Carnegie UK said in its June 2019 submission to the Online Harms White Paper consultation:

“Worryingly, there are references to proactive action in relation to a number of forms of content (and not just the very severe child sexual abuse and exploitation and terrorist content) which in the light of the emphasis in the codes could be taken to mean a requirement for upload filtering and general monitoring to support that.”

Demos’ submission to the draft Online Safety Bill Committee in September 2021 identified as a primary risk:

“A focus on regulation and moderation of content rather than platform systems which affect the risk of harm arising from that content” (emphasis in original)

and said:

“Although the Bill sets out a systems-based approach, there is a focus on reducing harm through content takedown measures, measuring the incidence of harms online and a focus on enforcing terms and conditions. ... we are concerned that in implementation this will turn into a ‘content-based approach’ by proxy, by prioritising the regulation of content moderation systems above other systems and design changes.”

 Demos' April 2022 position paper on the Online Safety Bill argued that:

“The Bill treats a ‘systems’ approach as meaning a ‘systems for dealing with content’ approach…”

The Demos position paper also expressed particular concern about the “strong risk of infringing on either privacy or freedom of expression” in Ofcom’s ability to require use of proactive content moderation technology.

The 5Rights Foundation’s response to Ofcom’s final Illegal Harms Code of Practice in December 2024 said:

“The legislation has a clear objective that services are made “safe by design” but the majority of Ofcom’s proposed measures are not designed to prevent harm occurring in the first place – instead focusing on content moderation and reporting tools. While greater requirements on governance and accountability are welcome, this in itself will not ensure safety by design.”

If content-focused measures, or at least automated filtering, are not a variety of safety by design then a definition of safety by design for insertion in the Act could be expected to exclude measures of that kind; albeit how it could do so when the Act specifically contemplates the imposition of automated content detection and filtering is a conundrum. 

But since the government officially regards automated content filtering as a safety by design measure, it would seem highly unlikely that a definition contradicting that could find its way into the Act.

Ofcom’s Online Safety Act implementation

With the closing of Ofcom’s Summer 2025 consultation on additional safety measures, we can assess how far the Code of Practice measures recommended or proposed by Ofcom to date are – and are not – focused on systems for content moderation.

The consultation in fact provides a dual opportunity: to analyse Ofcom’s existing and proposed measures from a safety by design perspective, and to look at how safety by design proponents have reacted to Ofcom’s newest proposals for automated content filtering (in Ofcom terminology, ‘proactive technologies’).

Non-content, reactive content-related and proactive content-related How far have non-content safety by design principles found expression in Ofcom’s implementation of the Online Safety Act?

Regardless of whether there is overlap between systems measures and content moderation measures, we can still conceive of functionality-oriented measures that do not require the platform to make judgements about content, nor involve directly limiting dissemination of content at all. A friction measure such as a warning ‘Did you mean to post without reading the linked article?’ would be an example of such a non-content measure.

Thus we can break down the measures so far recommended or proposed by Ofcom into non-content and content-related. The latter can be further divided into reactive and proactive.

In total, across the Illegal Content, Protection of Children and draft Additional Measures Codes for U2U services, there are (on my reckoning) 73 non-content measures, 27 reactive content-related measures and 12 proactive content-related measures. For illegal content, most of the proactive measures are contained in the Additional Measures consultation and are based on content detection and filtering technology of various kinds.

However, a closer look at the 73 non-content measures reveals that 50 of them are administrative, procedural or information provision: appointing an accountable individual, preparing various written documents, training, complaints and appeals procedures, publishing user support materials and so on. Whilst those are aspects of wider systems design, non-content measures addressed to features and functionality are of more immediate interest.

That leaves 23 non-content measures: 11 in the Illegal Content codes, all of which relate to children (in two cases only partially), and 12 in the Protection of Children codes.

Most of the 23 non-content measures concern technical functionality of the platform. The measures are limited (as required by the Act) to UK users and relate to:

• Implementing an age-assurance process (ICU B1, PCU B1)
• Use of highly effective age assurance (HEAA) (PCU B2 to B7) (Age assurance does of course indirectly affect the content available to users who are not verified as over-18, as the result of content-related measures predicated on age assurance.)
• Safety defaults for child users concerning connection lists, account recommendations and direct messaging (ICU F1)
• Removal of five kinds of functionality from child-user livestreams (ICU F3)
• Options for user account blocking, disabling comments (for child users, or in some circumstances all registered users) (ICU J1, ICU J2)
• Enabling children to give negative feedback on content recommender systems (PCU E3)
• Providing information to children, when they restrict content or interactions with other accounts, as to the effect of doing so and further options available (PCU F2)
• Options for user blocking and muting, disabling comments (users not determined to be adults by use of HEAA) (PCU J1, PCU J2)
• Positive consent to group chat invitations (users not determined to be adults by use of HEAA) (PCU J3)

These examples illustrate that non-content measures are feasible, albeit some of those measures are, at least in part, precursors to content-related measures. Most obviously, age assurance underpins not only some of the non-content measures listed above, but also measures about content that should be hidden from under-18s.

Generally, it is striking how many of Ofcom’s non-content functionality measures are concerned with denying functionality to, or to interactions with, under-18s.

As to content-based measures, the Additional Measures consultation marks a decided shift towards automated content detection. Should these be welcomed as a version of safety by design, deprecated as systems for content moderation, or regarded as a means of addressing residual issues that cannot be designed out?

Safety by design or ex-post? OSAN’s cross-cutting response to Ofcom’s Additional Measures consultation takes issue with Ofcom’s description of some content-related measures, including proactive technology, as being ‘safety by design’:

“While some of the proposed measures - including automated content moderation (para 1.51) and livestreaming (p27) - are framed by Ofcom as being “safer by design”, these are primarily about ex-post mitigations for harmful content (reporting content, or relying on user action after harm has occurred) or introducing a form of safety tech (proactive tech measures) rather than embedding safe design at the level of systems and processes. There is still no understanding of what good service redesign should look like to ensure a more holistic orientation towards safety.” (emphasis added)

However, OSAN’s companion detailed response to the Additional Measures Consultation characterises Ofcom’s proactive technology proposals as safety by design:

“We broadly support the move towards requiring proactive technology as a safety-by-design approach to user safety”.

The detailed response (but not the cross-cutting response) would therefore seem to endorse the government’s view of safety by design.

OSAN also suggested that Ofcom’s principles-based proactive technology proposals could be extended to include intimate image abuse.

Recommender systems The Demos Digital submission endorsed Ofcom’s proposed content-specific approach to recommender systems:

“The Demos Digital team agrees with Ofcom’s proposal to exclude illegal content from recommender systems until the content has been reviewed by content moderation teams.”

After pointing out that “Automated content identification tools are known to struggle with reliability and bias”, Demos Digital then suggested improvements including:

“Because of these risks of inconsistency, Ofcom should provide specific guidance for platforms’ responsible use of automated content identification tools, including: transparency reporting; quality control standards for automated identification systems, including bias, reliability and accuracy; impact assessments for evaluating the automated systems; and model parameters for identifying illegal content. We believe this would alleviate some of the risks of automated content identification systems – such as inconsistencies, inaccuracies, and bias – which could result in the over-exclusion of legal content, or under-exclusion of illegal content.”

At the level of principle it is difficult to see how this reflects a systems-based approach, other than in the sense of systems for moderating content.

Parenthetically, even if a tendency to bias could be alleviated, there is still the insoluble problem that automated content identification tools do not have access to off-platform contextual information that can affect legality of the user content in question.

In its comments on recommender systems OSAN supports limitations on the reach of “content that is harmful in nature”, if accompanied by freedom of expression safeguards such as explanations of how the systems work in practice, and notification of creators when their content is affected so as to allow them to use complaints and appeals processes.

Live-streaming For live-streaming, OSAN has suggested some concrete ways in which Ofcom’s proposed Additional Measures could go further: building in a delay to livestreaming and turning off livestreaming by default for under-18s or under-16s. It describes these as safety by design measures:

“15. Ofcom’s proposals focus on responding to harm after it occurs and content moderation rather than preventing it in the first place. There is no requirement for live-feed delays, which are standard practice in traditional broadcasting, to prevent harmful or illegal content from being aired in real time. Safety-by-design means including proactive measures such as time-delay buffers and real-time risk assessment. There is plenty of guidance available to broadcasters on this topic.” (emphasis added)

However, it then describes them as ex-post measures:

“17. More broadly, we would recommend that Ofcom consider a greater array of ex-post features - e.g. borrowing from broadcasting good practice and building more delay into a live stream as a feature.” (emphasis added)

Is time delay an example of safety by design or an ex-post feature? The distinction would not necessarily matter much, were it not for the fact that a statutory definition of safety by design is proposed. But either way, although a time delay is of itself a non-content measure, its purpose is to enable the platform to make judgements about the content being live-streamed and (if thought necessary) to shut down the stream. OSA describes that as real-time risk assessment. In the context of the Act, those would have to be judgements about illegality or (for child-accessible streams) content harmful to children. 

For children, OSAN contemplates a non-content-related measure: turning live-streaming off by default for children, whether under-16 or under-18. It also observes that “A strong understanding of safety-by-design would mean that where livestreaming cannot be delivered safely it shouldn’t be in place.”

Finally, OSAN cites Ofcom’s proposed limitation on livestream screen capture and recording for under-18s (part of ICU F3) as an example of friction.

Safety by design in context

As implementation of the Online Safety Act has progressed, it is perhaps not surprising if it has become less clear how safety by design should translate into concrete measures. The theory of online safety by design, founded on the notion of risk-creating features, was formulated in the context of a range of services and harms that differed greatly from those in scope of the Online Safety Act. The range of services within the Act is far broader and the kinds of harm are much more specific.

In July 2018 Woods and Perrin, working with Carnegie UK, proposed a:

“Virtuous circle of harm reduction on social media. Repeat this cycle in perpetuity or until behaviours have fundamentally changed and harm is designed out.” (Harm Reduction in Social Media, 17 July 2018 ) 

As to kinds of services, the proposal was aimed at around 10 social media companies each with over 1 million users. By January 2019, after discussion with various stakeholders, the authors had decided to extend the proposal to cover ‘social media and other internet platforms’ regardless of size. Now the Act covers an estimated 25,000 UK services (100,000 or more worldwide), 80% of which are micro-businesses (less than 10 employees). 

On the face of it the underlying premise of the harm reduction cycle seems to be that what a user does on a platform is primarily the result of its design. However, the authors of the proposal say that their argument is not that we are 'pathetic dots' in the face of engineered determinism, but that the architecture of the platform nudges us towards certain behaviour (Woods and Perrin, Online harm reduction - a statutory duty of care and a regulator, April 2019.) 

Even if it can be said that algorithmically driven social media platforms nudge us towards certain behaviour, how would that apply outside that specific milieu, for instance to plain vanilla discussion forums? And if, even on those large social media platforms, design only nudges rather than determines user behaviour, how far can harm really be designed out of the system?

As to kinds of harms, the safety by design theory is premised on platforms being risk creators. We always then have to ask, risk of what? In the context of the Online Safety Act that means connecting a given feature to a created or exacerbated risk of one of the specific kinds of criminality in scope of the Act, or of specific kinds of content harmful to children.

Within the context of the Act, the theory has never been easy to render into concrete expression:

• If the idea is that a user’s decision to post, say, an illegal offer to ferry illegal immigrants across the Channel is down to the design of the platform, that seems implausible.

• If the idea is that platform design can prevent such content being encountered, but without descending into content moderation and filtering, how is that to be done? Similarly if the concern is to prevent specific kinds of content being repeated or stimulated.

• If it means that recommender algorithms could be designed in ways that lessen the likelihood of their disseminating illegal content, it would have to be explained how that can be achieved without trespassing into content filtering.

• If the idea is that platform functionality can be designed to make it harder or slower to post, share or comment on user content generally, or to impose volume limits (a ‘circuit-breaker’), that would fit the theory. However, that kind of friction measure would necessarily strike against desirable and undesirable content alike, raising human rights proportionality issues.

• If the idea is that some functionalities should be banned, that would fit a version of the theory that holds that some functionalities cannot be designed safely. But the more general purpose the functionality in question, the greater the impact on legitimate content and the greater the human rights challenge.

• If the idea is that harm to children can be prevented by platform design which, for instance, reduces opportunities for adults to contact children, that would fit the theory.

If no connection can be found between a given technical or business model feature of a platform and a risk of a user deciding to behave illegally in a particular way, then the regulator will look somewhere other than those design features to counter illegality: to other design features or, failing that, to systems for moderation. 

Professor Woods has suggested that designers should ask themselves: ‘What happens when the bad people get hold of this feature?’ (Introducing the Systems Approach and the Statutory Duty of Care, ibid.) However, that question could be asked of any general purpose functionality, risk-creating or not. On the face of it the question is about possible uses, not whether the feature in question creates or exacerbates a risk of a particular illegal or harmful use. It could be asked of the very act of providing a forum to which users can post. If we are not careful, we rapidly fall into the trap of characterising speech as a risk, not a fundamental right.

It is telling that Ofcom adopted that same approach in its statutory Risk Register: rather than attempt to identify functionalities that inherently create or exacerbate risk of illegality or content harmful to children, it sought to identify features that are used by malefactors as well as by law-abiding users: correlation rather than causation. That led it to list as risk factors general purpose functionality such as the ability to create hyperlinks.

If safety by design turns out to be a poor fit with much of the Online Safety Act, it should be acknowledged that the originators of the safety by design theory never wanted illegality to be the touchstone in the first place. Professor Woods said:

“These categories of harm should be identified by reference to their impact on the victim, not by reference to whether the speech might be considered illegal or not.” (Introducing the Systems Approach and the Statutory Duty of Care, ibid.)

That risks a leap from the frying pan (attributing risk of illegal behaviour to a platform feature) into the fire (pursuing nebulous and subjective kinds of harm). That aside, it would be no surprise if the theory turns out not to map easily on to the Act. It is one thing to say that, for instance, chasing ‘Likes’ trains users to produce ‘response-creating content’ (Introducing the Systems Approach and the Statutory Duty of Care, ibid). It is something else to show that a feature creates a risk of a user committing a specific criminal offence.

It may not be fanciful to think that something has got lost along the way from the 10 or so large social media platforms that the Carnegie UK authors had in mind for their original 2018 proposals, to the broad variety of 100,000 UK and overseas services in scope of the Online Safety Act. If, in essence, the theory was always really about large social media companies, their curation and engagement algorithms and their data-driven business models, it would not be a shock to find that it turns out to have little or no application beyond that.

For platforms where user agency is the predominant factor, and design decisions cannot realistically be regarded as likely to increase or decrease the likelihood of illegality or relevant content harm, logic would suggest that issues that cannot be designed out would most likely be at the forefront, not residual. A fruitless quest for specific illegality- or harm-inducing features could then easily result in a theoretical focus on systems and processes lapsing into systems for content moderation, thence to proactive content filtering technologies.

As to a statutory definition of safety by design, if systems for content moderation, including automated content filtering, are now to some extent embraced as an aspect of safety by design, it is difficult to see how a corresponding statutory definition could place meaningful limits on the kinds of concrete measures contemplated. It would also seem to have moved a very long way from the original conception of safety by design. 

If the reality is that we do not have a clear idea of how safety by design is meant to translate into concrete regulatory measures within the context of the Act, that would not be a good starting point for crafting a statutory definition.

The alternative, of course, is that I have always had safety by design wrong and that Parliament knew exactly what it intended in Section 1. If so, mea culpa.

Extraterritoriality and the transatlantic free speech wars

The transatlantic free speech wars continue to rage. The US House Judiciary Committee was in action again last week, taking aim at the European Commission (who rejected its latest interim report as ‘pure nonsense’) and provoking EU civil society groups in the process.

The US administration, for its part, fired off its most recent salvo shortly before Christmas last year, when US Secretary of State Marco Rubio added five people to a list of individuals who would not be allowed visas, due to their activities in the 'global censorship-industrial complex'.

The US rogues' gallery included former EU Commissioner Thierry Breton, whose letter to Elon Musk in August 2024, referencing the Digital Services Act, scuppered Breton's prospects of a job in the 2024-2029 European Commission. The Commission have been trying to live down the letter ever since. US critics of the DSA have never let them forget it.

Notably, or perhaps prudently, the no-visa list included no current foreign state officeholders or functionaries. It did not go as far as when the US imposed visa restrictions on the Brazilian Supreme Court judge Alexandre de Moraes in July 2025. The implied threat, however, remains: "The State Department stands ready and willing to expand today's list if other foreign actors do not reverse course."

The next, heavily trailed, US counterstrike may be legislative: a federal ‘GRANITE Act’ Bill. We await to see if such a Bill materialises, and if so what it consists of. A state-level GRANITE Act was introduced into the Wyoming legislature yesterday.

If a federal Bill were framed along the lines of the 2010 SPEECH Act (aimed at libel forum-shopping) it would act as a shield, explicitly preventing enforcement of foreign regulatory and similar orders within the USA. A more radical (and controversial) step would be if it contained a sword: a cause of action on which aggrieved plaintiffs could claim damages in the US courts. The most controversial step would be if that were accompanied by amendment of the US Foreign Sovereign Immunities Act to enable foreign regulators such as Ofcom to be sued, either in the federal courts or under state legislation such as the Wyoming Bill.

Sovereignty exercised or violated?

What exactly is the US administration aggrieved about? Secretary of State Rubio's social media announcement referred to:

"egregious acts of extraterritorial censorship" by “ideologues in Europe [who] have led organized efforts to coerce American platforms to punish American viewpoints they oppose.”

The official State Department statement added:

“These radical activists and weaponized NGOs have advanced censorship crackdowns by foreign states—in each case targeting American speakers and American companies.”

It went on:

“President Trump has been clear that his America First foreign policy rejects violations of American sovereignty. Extraterritorial overreach by foreign censors targeting American speech is no exception.”

Stripped of the rhetoric, this is at least in part an accusation that the EU and UK, implicitly breaching international law on territorial sovereignty, have overreached in asserting their local regulatory regimes across the Atlantic.

The European Commission's response to the US visa bans asserted the EU's own:

"sovereign right to regulate economic activity in line with our democratic values and international commitments .... If needed, we will respond swiftly and decisively to defend our regulatory autonomy against unjustified measures."

The reported UK response was more anodyne:

"While every country has the right to set its own visa rules, we support the laws and institutions which are working to keep the internet free from the most harmful content."

Neither response directly addressed the US’s complaint about extraterritoriality. Since Secretary of State Rubio’s announcement did not identify any specific foreign state act that had prompted the visa sanctions, that was perhaps unsurprising. Moreover the visa restrictions were aimed, Thierry Bretton apart, at private persons who had not held public office. As against them, the US complaint was of ‘advancing’ censorship crackdowns by foreign states.

The French Foreign Minister, for his part, claimed that the DSA:

“…has absolutely no extraterritorial reach and in no way affects the United States.” (Jean-Noël Barrot, tweet, 23 Dec 2025)

Taking sides

What should a dispassionate legal observer make of all this? Some, no doubt, will be tempted just to plump for one side or the other, motivated by partisan preference for the EU, UK or US approach to governing speech and online platforms, by broader political affinities, or by views on the propriety or otherwise of deploying visa sanctions for this kind of purpose.

Tempting as that may be, simply to declare 'four legs good, two legs bad' will not do when it comes to considering international law rules and extraterritoriality. Taking sides based purely on a preference for the Digital Services Act or the Online Safety Act over the US First Amendment, or vice versa, does not address the underlying legal issue: how, in the inherently cross-border online world, to go about drawing boundaries - or at least minimise friction - between different national or regional legal systems. A more analytical approach is called for.

Prescriptive versus enforcement jurisdiction

For that we have first to distinguish between prescriptive and enforcement jurisdiction. Prescriptive jurisdiction is the territorial ambit of legislation: how far, and on what basis, does it claim to apply to persons or conduct outside its borders? Enforcement jurisdiction, on the other hand, is about concrete exercise of powers by a state authority. In the case of the Online Safety Act that authority is the designated regulator, Ofcom.

Where extraterritoriality is concerned, international law gives more leeway to prescriptive than to enforcement jurisdiction. That is because the state’s conduct in legislating is merely assertive. Although laws are an expression of state power, writing something into a state’s own legislation does not of itself involve conduct on the territory of another state.

Typically it is enforcement that causes problems, both in its own right - steps that the authorities have taken, especially cross-border, to enforce against a foreign person - and in the light that enforcement shines on the prescriptive territorial reach of the substantive legislation.

The US complaint – prescriptive or enforcement?

Did the US complaint concern prescriptive or enforcement jurisdiction? An interview given by Under-Secretary of State Sarah B. Rogers to the Liz Truss Show before Christmas put a little more flesh on the bones as far as the Online Safety Act is concerned. She suggested that European, UK and other governments abroad were trying to nullify the American First Amendment and that: 

"when British regulators decree that British law applies to American speech on American sites on American soil with no connection to Britain, then we're kind of forced to have this conversation." 

She went on:

"The position that Ofcom has taken in the 4Chan litigation is essentially that I, an American, could go set up a website in my garage, it could be Sarah's hobby forum, it could be all about America, it could be all about the 4th of July or whatever. It could have no employees in Britain, no buildings in Britain, my speech wouldn't even need to reach into Britain. I'm not posting about the Queen or anything, I'm posting about American concepts, American political controversies. Ofcom's legal position nonetheless is that if I run afoul of British content laws, then I have to pay money to the British government. When that happens, I think you should expect a response from the American government, and I expect to see one shortly." 

On the face of it Rogers’ concern is about the substantive territorial ambit of the OSA: in other words, prescriptive jurisdiction. 

Prescriptive jurisdiction

International law recognises various grounds on which extraterritorial prescriptive jurisdiction can be regarded as justified, some of them potentially very broad. These tend to reflect a broader principle that there must be sufficient connection between the person or conduct and the state asserting jurisdiction to justify the extraterritoriality in question. The more tenuous the connection and the greater the cross-border reach, the more exorbitant the claim to jurisdiction and the less likely that the extraterritoriality can be justified. 

That is the theory. In practice, the customary norms of international law tend to be distinctly malleable and, when push comes to shove, to merge into geopolitics.

Enforcement jurisdiction

In contrast, for exercise of investigative or enforcement jurisdiction, the traditional view is that nothing less than consent of the target state will do. Unlike for prescriptive jurisdiction, there is no balancing exercise to justify the degree of extraterritoriality of the asserted jurisdiction. The focus is entirely on the conduct of the state and whether it is an incursion on the territorial sovereignty of the target state.

However, this principle has come under strain. When electronic communication and the internet enable state authorities to act remotely without setting foot in another state’s territory or sending a physical document across the border, does that violate another state’s territorial sovereignty? Should, as for prescriptive jurisdiction, other factors come into play that could justify the state's conduct?

For one answer we can go back to 1648 and the Peace of Westphalia. This was the birth of the modern nation state, in which each state has exclusive sovereignty over its own territory. The corollary of that principle is an aversion to projection of state power into another state's territory: most obviously, sending troops across the border.

That, however, is not the only way of violating a state's sovereignty. Enforcement actions such as serving a court order or an arrest warrant within a foreign state's territory also project state power across the border and are considered to require the consent of the nation state concerned:

"Persons may not be arrested, a summons may not be served, police or tax investigations may not be mounted, and orders for production of documents may not be executed on the territory of another state, except under the terms of a treaty or other consent given." (Brownlie's Principles of Public International Law (9th edn) J. Crawford, Oxford, 2019. p.462)

That is why there is a proliferation of international treaties dealing with issues such as cross-border service of legal proceedings, assistance from overseas authorities in obtaining evidence for criminal prosecutions (MLAT) or, more recently, enabling direct service of information requests on foreign telecommunications operators. 

Enforcement jurisdiction and regulators

A requirement for consent of the target state creates a potential problem for regulators, whose procedures are highly bureaucratic: inevitably so since considerations of due process and fundamental rights will require them to give enforcement targets full and fair notice of their proposed and actual decisions. They are also often given powers to serve mandatory demands for information, backed up by sanctions (sometimes criminal offences, sometimes civil penalties).

On what basis can a regulator send such official documents across borders without impinging on the sovereignty of the target state? The answer is not immediately obvious, especially since the activities of regulators do not necessarily fall into simple categories of civil or criminal upon which international treaties regarding service of legal documents tend to be founded.

A typical solution to the territorial sovereignty problem is to enlist the assistance of the relevant authorities in the target state. If such assistance is not covered by a multinational or bilateral treaty, a regulator might come to an arrangement such as a memorandum of understanding between agencies in a group of states.

The 2020 multilateral Competition Authorities Mutual Assistance Framework model agreement, for instance, envisages that requests for voluntary provision of information could be made by a direct approach to persons in another territory. For mandatory process the route is via the authorities in the other country.

However, courts have sometimes held that serving a cross-border notice is not like trespassing on the territory of the receiving state. In the UK the Court of Appeal in Jimenez considered an HMRC taxpayer information notice (with the potential sanction of civil, but not criminal, financial penalties) served by post on someone in Dubai, in order to check his UK tax position. Jimenez argued that sending the notice was contrary to international law, as it would:

“offend state sovereignty by violating the principle that a state must not enforce its laws on the territory of another state without that other state’s consent.”

Leggatt LJ (as he then was) said:

“I do not accept that sending a notice by post to a person in a foreign state requiring him to produce information that is reasonably required for the purpose of checking his tax position in the UK violates the principle of state sovereignty. Such a measure does not involve the performance of any official act within the territory of another state – as would, for example, sending an officer of Revenue and Customs to enter the person’s business premises in a foreign state and inspect business documents that are on the premises…”.

The Jimenez decision postdated the current edition of Brownlie quoted above. In KBR the UK Supreme Court emphasised that Jimenez concerned civil, not criminal, penalties. 

All this is not to say that a domestic statute can never expressly grant powers to take steps that, as a matter of enforcement jurisdiction, could go further than envisaged by international law. A UK statute may indeed do that, but in the expectation that the powers will be used with restraint, in a way that does not offend the sensibilities of another nation state (a.k.a. comity).

This passage from the Court of Appeal judgment in Competition and Markets Authority v Volkswagen and BMW, a case upholding CMA information notices served on German companies, is illuminating:

“All competition authorities worldwide face the same conundrum. Their statutory duty is to preserve the integrity of their domestic markets and protect consumers; yet, to perform that task regulators frequently have to focus their fire power upon actors located abroad where, if they seek enforcement, they might confront a variety of legal and practical problems. … How do legislatures square the circle? They achieve this by conferring broad extraterritorial regulatory and investigatory powers which can be exercised in undiluted form within their territorial jurisdictions, but which are exercised with circumspection and pragmatism when dealing with undertakings physically located elsewhere.”

The judgment went on to discuss comity:

“The creation of a power to be exercised with comity in mind … is, in our judgment, an eminently apt device to enable regulators to address, flexibly, issues of comity if and when they arise. [Counsel] for the CMA explained how comity worked. She acknowledged candidly that, notwithstanding the existence of broad investigatory powers, it was ‘out of the question’ that the CMA would for instance ever seek to conduct an on the spot investigation (a dawn raid) at the premises of an undertaking physically located outside the jurisdiction. Equally, she did not shirk from acknowledging that there could be difficulties in the exercise of mandatory powers of enforcement or sanction against a foreign undertaking which failed to comply with a statutory request for information. Such practical difficulties were simply the stuff of a regulator’s life.” 

Thus from a comity perspective a national regulator seeking to take enforcement steps against a foreign person may decide to tread carefully, especially where the subject matter may touch on particular sensitivities of the target state, even if the domestic legislation gives it power to act across borders. A combination of ambitiously extraterritorial prescriptive jurisdiction and broad investigatory and enforcement powers has the potential to become a combustible mixture.

Online Safety Act – prescriptive jurisdiction

With that background out of the way, how do Under-Secretary Rogers’ comments stack up?

In terms of its overall ambit, the UK Online Safety Act is extraterritorial but does not go as far as the mere accessibility position taken by Australian online safety legislation. The Australian Online Safety Act 2021 baldly asserts that a social media service is in scope of the Act unless “none of the material on the service is accessible to, or delivered to, one or more end-users in Australia”.

That legislation gave rise to civil litigation for an injunction brought by the eSafety Commissioner in the Australian courts, arguing that X should be required to take down certain videos worldwide and that geofencing to exclude Australia was insufficient. The regulator lost. 

The UK OSA sets out three grounds on which a service can be regarded as ‘UK-linked’ and so be a regulated service within scope of the Act. In the US litigation brought by 4Chan, Ofcom relies on two of those grounds: first, a significant number of UK users (based on statistics gleaned from 4Chan's website), and second the UK as a target market of the site (based on 4Chan seeking advertisers by reference to the percentage of UK users stated on its website).

Ofcom's position is thus that a sufficient UK connection (as stipulated in the OSA) exists in order for the OSA to apply to 4Chan. Ofcom did not (and could not) assert that the OSA safety duties apply to a site regardless of whether it has any UK connection.

How, then, should we interpret Under-Secretary Rogers' comment: “…when British regulators decree that British law applies to American speech on American sites on American soil with no connection to Britain”? That must presumably reflect a view of what should constitute a UK connection that is at odds with the OSA's criteria, or perhaps with Ofcom's interpretation of those criteria. It cannot, however, be argued that the OSA contains no UK connection criteria at all.

As to compliance with international law, the UK government would no doubt argue that as a matter of prescriptive jurisdiction the criteria for UK links stipulated in the OSA provide a sufficiently close connection with the UK to justify bringing a foreign service provider into scope, and are not exorbitant. It would also no doubt point to the fact that the substantive measures that can be required of an in-scope provider apply only to UK users of the service.

The Online Safety Act UK links criteria

The UK links set out in the OSA vary in the closeness of the stipulated UK connection. Some of them could be regarded as overreaching.  

The first ground - a 'significant' number of UK users - suffers from the vagueness of the term 'significant'. The Act does not elaborate on what that might mean and Ofcom has avoided specifics in its published guidance.

Speaking for myself, I have long proposed that extraterritorial jurisdiction on the internet should depend on whether a foreign site has engaged in positive conduct towards the jurisdiction. On that basis a self-contained test based only on number or proportion of users is potentially problematic: users may come to a site in numbers without the site operator ever having engaged in any positive conduct towards the country in which they are located. (This criticism can be applied to the DSA as well as the OSA).

The second ground - the UK as a target market - is reasonably conventional, if interpreted so as to reflect a requirement for positive conduct. Directing and targeting of activities has long been thought to be an appropriate ground on which to assert jurisdiction over internet actors.

The third ground - 'material risk of significant [physical or psychological] harm " is the most far-reaching and comes closest to a ‘mere accessibility’ test.

So far as prescriptive jurisdiction is concerned then, the OSA does stretch the limits of extraterritoriality, but – unless one were to take the position that a site is connected only to the country of its location - does not purport to apply to sites regardless of whether they have any connection to the UK.

Online Safety Act – enforcement jurisdiction

Although Under-Secretary Rogers’ comments are framed in terms of the OSA’s prescriptive jurisdiction, the point that 4Chan has emphasised in its US litigation concerns Ofcom’s exercise of its enforcement jurisdiction – serving a series of documents, including a mandatory information request under Section 100 of the OSA, directly on 4Chan by email.

As already noted, a regulator such as Ofcom proceeds against a service provider by way of a series of official notices. These present no jurisdictional problems if they can be served within the UK, but can Ofcom serve a notice on a foreign operator without violating the territorial sovereignty of its host country? As a matter of UK domestic law the OSA provides a variety of methods of service, including cross-border service by post and service by email.

Some might suggest that that, as a matter of international law, is an impermissible exercise of enforcement or investigatory jurisdiction unless done with the consent of the USA. But as Jimenez illustrates, a UK court would not necessarily agree; and in any case, if the words of the domestic statute are sufficiently clear to rebut any interpretative presumption against extraterritoriality, a UK court will give effect to them. 

Consequences

The consequences of exceeding acceptable limits of extraterritoriality may vary widely, depending on how exorbitant is the exercise of jurisdiction and how sensitive is the subject matter. They range from no response (often the case for merely prescriptive jurisdiction), to diplomatic, to refusal by foreign courts to recognise or enforce, to enactment of various kinds of blocking legislation.

One example of the latter was the US SPEECH Act 2010, which prevents enforcement of certain foreign libel judgments in the US courts and enables US persons to start proceedings for a declaration of non-enforceability in the US courts.

Another was the UK Protection of Trading Interests Act 1980. This was a response to a long period of US anti-trust legislation being enforced against conduct outside the USA by non-US companies. Years of diplomatic activity had failed to resolve the conflict, which had become more acute in the late 1970s.

As already mentioned, a state-level GRANITE Act has been introduced as a Bill into the Wyoming legislature. It is both a shield and a sword, albeit that the sword would be dependent on a federal amendment to the Foreign Sovereign Immunities Act. We await to see if a federal GRANITE Act will materialise. 

The Court of Appeal in the BMW case noted that it had been said that antitrust law was the best illustration of the problem of national public interest risking conflicting with issues of international sovereignty. Speech on the internet – today, online safety in particular - is bidding fair to seize that mantle.

Repeal, reform, rewrite?

A Parliamentary petition calling for repeal of the Online Safety Act has reached over 550,000 signatures and is due to be debated on 15 December.

The demand for abolition is eye-catching, but inevitably lacks nuance. In any case it is tolerably clear that the petition is aimed not at the entire Act, but at its core: the set of regulatory safety duties imposed on platforms and search engines. Even for those elements, the petition envisages not bare repeal but repeal and replacement:

“We believe that the scope of the Online Safety act is far broader and restrictive than is necessary in a free society.

For instance, the definitions in Part 2 covers online hobby forums, which we think do not have the resource to comply with the act and so are shutting down instead.

We think that Parliament should repeal the act and work towards producing proportionate legislation rather than risking clamping down on civil society talking about trains, football, video games or even hamsters because it can't deal with individual bad faith actors.”

Of course the Act contains much more than the safety duties: in particular, new communications and other criminal offences, age assurance requirements on non-user-to-user pornography websites in Part 5, and separate duties around fraudulent paid-for advertising.

Communications offences

Most, if not all, the new criminal offences either improve on what went before or are eminently justifiable. Is anyone suggesting, for instance, that deliberate epilepsy trolling or cyberflashing should not be an offence?

Possibly the only offence about which there could be some debate is the S.179 false communications offence. It has, rightly or wrongly, become something of a lightning rod for critics concerned about over-broadly criminalising disinformation.

Whatever doubts may exist about its current formulation, the S.179 offence is still an improvement on what went before. Its antecedents can be traced back to 1935, when distressing hoax telegrams were a concern:

“there are a number of cases where people … will send a telegram to a person saying that somebody is seriously ill and the recipient is put to great anxiety, sometimes even to expense, and then finds it is a bogus message. … That silly practice which causes anxiety - in some cases a telegram has been sent declaring a person to be dead when there was no foundation for the statement - ought to be stopped.” (Postmaster-General, Hansard, 5 March 1935)

The 1935 offence covered telegrams and telephone calls. In 1969 it was broadened to include messages sent by means of a public telecommunications service. In 2003 it was amended to cover communications sent across a public electronic communications network. By that time internet communications were undoubtedly in scope. Following a Law Commission recommendation, S.179 OSA replaced the 2003 offence.

The 1935 (and 1969 and 2003 updated) offence was framed in terms of causing “annoyance, inconvenience, or needless anxiety”. S.179’s “non-trivial physical or psychological harm”, whatever the criticism that it has attracted as a definition, is narrower.

As to the application of S.179 to fake news, the Law Commission said:

“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.” (Modernising Communications Offences, Final Report, 20 July 2021, para 3.3)

When the new offences came into force in January 2024 the government announced that the false information offence criminalised “sending fake news that aims to cause non-trivial physical or psychological harm.”

Notwithstanding the improvement on its predecessors, S.179 shares some features in common with the harmful communications offence (also recommended by the Law Commission) which was rightly dropped from the Bill. As a result of that decision the long-criticised S.127(1) Communications Act 2003 and S.1(a)(i) Malicious Communications Act 1988 “grossly offensive” offences remain in force. A re-examination of how to replace those is overdue. In conjunction, another look could be taken at the false information offence.

The core safety duties

Returning to the core safety duties, do they merit repeal? The underlying point of principle is that the Act’s safety duties rest on defectively designed foundations: a flawed analogy with the duty of care owed to visitors by the occupier of real world premises. That, it might be said, requires exposing the root cause, not compilation of a snagging list.

Why is the analogy flawed? The Act’s safety duties are about illegal (criminal) user content and activities, and user content harmful to children. Occupiers’ liability is about risk of causing physical injury. Those are categorically different. You don’t need 444 pages of Ofcom Illegal Content Judgements Guidance to decide that a projecting nail in a floorboard poses a risk of physical injury. Moreover the remedy – hammering down the nail — hurts no-one.

Content judgements about illegality, in particular, tend to be nuanced, complex and will often depend on contextual information that is unavailable to the platform. Hammering down nails at scale inevitably results in arbitrary judgements and suppression of legitimate user content. That is magnified when the Act contemplates automated content detection and removal in order to fulfil a platform’s duties.

In short, speech is not a tripping hazard and it was folly to design legislation as if it were.

Nor, it might be suggested, should anyone be thinking of adding more layers to this already teetering wedding cake. It is a confection with which no-one – supportive or critical – is currently very happy, albeit for widely differing reasons. If vaguer and more subjective notions of harmful content are piled onto the legislation, the structural cracks can only expand.

So, where may critics’ attentions be focused? The Open Rights Group, both itself and jointly with other civil society organisations, has produced detailed briefing papers ahead of the Parliamentary debate. 

Here, using U2U services by way of illustration, are some thoughts of my own on possible areas of interest.

S.10(2)(a) Proactive content filtering

Section 10(2)(a) imposes a duty on platforms to take or use proportionate measures relating to the design or operation of the service to “prevent individuals from encountering priority illegal content by means of the service”. Priority illegal content is a list of over 140 criminal offences, plus their inchoate versions: encouraging and assisting, conspiracy and so on.

The EU Digital Services Act adopts a diametrically opposed stance, retaining the long-standing EU prohibition on imposing general monitoring obligations on platforms.

A platform that complies with recommendations made in an Ofcom Code of Practice is deemed to comply with the duty laid down in the Act. Ofcom started, in its original Illegal Content Code of Practice, by recommending CSAM perceptual hash-matching and URL-matching for some platforms.

It is now going further in its Additional Safety Measures consultation, proposing to apply perceptual hash-matching to terrorism and intimate image abuse content, and proposing (among other things) ‘principles-based’ proactive technology measures for a broader range of illegal content. Unlike the previous CSAM hash-matching recommendation, IIA perceptual hash-matching could be carried out against an unverified database of hashes.

However, once extended beyond matching against a list of verified illegal items, the assumption underpinning content detection and upload filtering – that technology can make accurate content judgements – becomes questionable. Automated illegality judgements made in the absence of full contextual information are inherently arbitrary. Measures that result in too many false positives cannot be proportionate.

Ofcom’s proposed ‘principles-based’ measures are its most controversial recommendations. Ofcom has declined to specify what is an acceptable level of false positives, leaving that judgement to service providers. As Ofcom’s Director for Online Safety Strategy Delivery Mark Bunting said in evidence to the Lords Communications and Digital Committee:

“We think it is right that firms have to take responsibility for making the judgment themselves about whether the tool is sufficiently accurate and effective for their use. That is their judgment, and they should use it; we are expecting that to drive wider use than the regulator just issuing directions.” (14 October 2025)

Ofcom itself acknowledges that its principles-based proposals:

“could lead to significant variation in impact on users’ freedom of expression between services”. (Consultation, para 9.136)

That conflicts with the reasonable degree of certainty for users necessary for compliance with the ECHR.

Section 10(2)(a) is likely to be high up the critics’ list: too general, flawed in its underlying assumptions, and (as is now evident), a means for Ofcom to propose principles-based proactive technology measures that are so vague as to be non-compliant with basic rule of law and ECHR requirements.

S.192 Illegality judgements

A related issue, but relevant also to reactive content moderation, is the standard that the Act sets for illegality judgements. S. 192 specifies how service providers should go about determining whether a given item of content is illegal (or, for that matter, whether it is content of another relevant kind such as content harmful to children).  The provision was introduced into the Bill by amendment, following the Independent Reviewer of Terrorism Legislation’s ‘Missing Pieces’ paper which rightly pointed out the Bill’s lack of clarity about how service providers were to go about adjudging illegality, especially the mental element of any offence.

Section 192, although it clarifies the illegality judgement thresholds, bakes in removal of legal content: most obviously in the stipulated threshold of “reasonable grounds to infer” illegality. It also requires the platform to ignore the possibility of a defence unless it has positive grounds to infer that a defence may succeed.

The S.192 threshold also sits uneasily with S.10(3)(b), which requires a proportionate process for swift removal when a platform is alerted by a person to the presence of any illegal content, or “becomes aware” of it in any other way.

S.121 – Accredited technology notices

This is perhaps the most overtly controversial Ofcom power in the Act. For CSAM, the S.121 power can – unlike the preventive content duty under S.10(2) – require accredited proactive detection technology to be applied to private communications. Critics have questioned whether the power could be used to undermine, circumvent or render it impossible for a service provider to use end to end encryption.

As with proactive technology measures, in its consultation on minimum standards of accuracy of accredited technology for S.121 purposes Ofcom has declined to specify quantitative limits on what might be an acceptable level of accuracy of the technology.

Ofcom’s relationship with government.

The extent to which government is in a position to put pressure on Ofcom, in principle an independent regulator, has come into clearer focus since Royal Assent. The most notable example is the Secretary of State’s ‘deep disappointment’ 12 November 2025 letter to Ofcom.

The Act also contains specific mechanisms giving the Executive a degree of control, or at least influence, over Ofcom.

S.44 Powers of direction These powers enable the Secretary of State to direct OFCOM to modify a draft code of practice on various grounds, including national security, public safety, public health, or relations with a foreign government. The original provisions were criticised during the passage of the Bill and were narrowed as a result. So far they have not been used.

S.92 and S.172 Strategic priorities S.172 enables the Secretary of State to designate a statement setting out the government’s online safety strategic priorities. If it does so, OFCOM must have regard to the statement when carrying out its online safety functions. The government has already used this power.

Whatever Parliament may have thought was meant by ‘strategic’ priorities, or ‘particular outcomes identified with a view to achieving the strategic priorities’, the 30 page Statement designated by the Secretary of State on 2 July 2025 goes into what might be thought to be operational areas: for instance, interpreting ‘safe by design’ as contemplating the use of technology in content moderation.

S.175 Special circumstances directions The Secretary of State has power to give a direction to Ofcom, in exercising its media literacy functions, if the Secretary of State has reasonable grounds for believing that there is a threat to the health or safety of the public, or to national security. A direction could require Ofcom to give priority to specified objectives, or require Ofcom to give notice to service providers requiring them to make a public statement about steps they are taking in response to the threat. The grounding of this power in Ofcom’s media literacy functions renders its ambit somewhat opaque.

S.98 Ofcom’s risk register and sectoral risk profiles

S.98 mandates Ofcom to produce risk registers and risk profiles for different kinds of services, grouped as Ofcom thinks fit according to their characteristics and risk levels. ‘Characteristics’ includes functionalities, user base, business model, governance and other systems and processes.  ‘Risk’ means risk of physical or psychological harm presented by illegal content or activity, or by content harmful to children.

In preparing its work product Ofcom abandoned any notion that functionality has to create or exacerbate a risk of illegal content or offences. Instead, Ofcom’s risk register is based on correlation: evidence that malefactors have made use of functionality available on platforms and search engines.

That has led Ofcom to designate common or garden functionality, such as the ability to use hyperlinks, as risk factors. That, it might be thought, turns the right of freedom of expression on its head. We do not treat the ability to use pen and paper, a typewriter, or a printing press as an inherent risk.

The length, complexity and often impenetrability of Ofcom’s work product is also noteworthy. The Illegal Harms Register of Risks runs to 480 pages, accompanied by 84 pages of Risk Assessment Guidance and Profiles.

Harm, illegality, or both?

The Act’s safety duties vary as to whether they are trying to protect users from risk of encountering illegal content per se, risk of suffering harm (physical or psychological) as a result of encountering illegal content, or both. This may seem a rather technical point, but is symptomatic of the Act’s deeper confusion about what it is trying to achieve.

Ofcom’s efforts to simplify the duties by referring to ‘illegal harms’ in some of their documents added to the confusion. Nor were matters helped by the Act’s designation, as priority offences, of some offences for which the likelihood of physical or psychological harm would appear to be remote (consider money-laundering, for instance).

There is also a curious mismatch between what the Act requires for Ofcom’s Risk Registers and Risk Profiles, compared with risk assessments carried out by service providers. Ofcom’s work products are required only to consider risk of harm (physical or psychological) presented by illegal content, whereas service provider risk assessments are also required to consider illegality per se.

S.1 The purpose clause

Section 1 of the Act is unlikely to be in the sights of many critics. But, innocuous as it may appear, it deserves to be.

The purpose clause ostensibly sets out the overall purposes of the Act. It was the last minute product of a new-found spirit of cross-party collaboration that infused the House of Lords in the final days of the Bill. In reality, however, it illustrates the underlying lack of clarity about what the Act is trying to achieve.

Purpose clauses are of debatable benefit at the best of times: if they add nothing to the text of the Act, they are superfluous. If they differ from the text of the Act, they are prone to increase the difficulty of interpretation. This section uses terminology that appears nowhere else in the Act, and is caveated with ‘among other things’ and ‘in broad terms’.

Possibly the low point of Section 1 is the reference to the need for services to be ‘safe by design’. Neither ‘safe’, nor ‘safe by design’ are defined in the Act. They are susceptible of any number of interpretations.

One school of thought regards safety by design as being about giving thought at the design stage to safety features that are preferably content-agnostic and not focused on content moderation. The government, in its Statement of Strategic Priorities, takes a different view: safety by design is about preventing harm from occurring in the first place. That includes deploying technology to improve the scale and effectiveness of content moderation.

That interpretation readily translates into proactive content detection and filtering technology (see the discussion of section 10(2)(a) above). Indeed Ofcom, in its response to the government’s Statement of Strategic Priorities section on safety by design, refers to its own consultation on proactive technologies.

Fundamental rethink?

There are more problems with the Act: vague core definitions that even Ofcom will not give a view on, over-reaching territoriality, the inclusion of small, low risk volunteer-led forums, questions around age assurance and age-gating, concerns that the definitions of content legal but harmful to children are imprecise and may deny children access to beneficial content, and others.

The most radical option would be to rethink the broadcast-style ‘regulation by regulator’ model altogether. This observer has always viewed the adoption of that model as a fundamental error. Nothing that has occurred since has changed that view. If anything, it has been reinforced. Delay, expense and an inevitably bureaucratic approach were hard-wired into the legislation. The opportunity cost of the years and resources spent heading down that rabbit hole has to be immense.

The results are now attracting criticism from all sides: supporters, opponents and government alike. The mystery is why everyone concerned could not see what was designed in to the legislation from the start. If you put your faith in a discretionary regulator rather than in clear legal rules, prepare for disappointment when the regulator does not do what you fondly imagined that it would. If you wanted the regulator to be bold and ambitious, be prepared for the project to end up in the courts when the regulator overreaches.

Finally, the Online Safety Act project has been bedevilled throughout by a tendency to equate all platforms with large, algorithmically driven, social media companies. Even now, a Lords amendment recently tabled to the Children’s Wellbeing and Schools Bill, claiming to be about “introducing regulations to prevent under 16s from accessing social media”, is drafted so as to apply to all regulated user-to-user services as defined in the Online Safety Act – a vastly wider cohort of services.

That takes us back to the flawed analogy with occupier’s liability. Possibly the analogy was conceived with large social media companies in mind. But then, if the projecting nail in the floorboard is actually a social media company’s engagement algorithm, not the user’s speech itself, that would suggest legislation based on a completely different foundation: one that focuses on features and functionalities that create or exacerbate a risk of specific, tightly defined, objectively ascertainable kinds of injury.

Put another way, if you want to legislate about safety, make it about safety properly so called; if you want to legislate about Big Tech and the Evil Algorithm, make it about that; if you want to legislate about children, make it about that.  

What alternative approaches might there be? I don’t pretend to have complete answers, but suggested some in my response to the Online Harms White Paper back in 2019; and again in this post, written during the 2022 hiatus while the Conservatives sorted out their leadership crisis.