Britain Draws a Line in the Sand: What the UK’s Under-16 Social Media Ban Means
The United Kingdom has announced that it will ban social media for children under the age of sixteen, with soon-to-be ex-Prime Minister Keir Starmer describing the measure as a line in the sand and a deliberate attempt to reset the relationship between a generation of children and the platforms that have shaped their childhoods. The government did not mince its words: tech giants, in the words of the official announcement, “had their chance and failed.”
This piece sets out what the UK is doing, how the policy works, how it departs from the regime that preceded it, and how it compares with approaches elsewhere. It draws on the early evidence now emerging from Australia, weighs the arguments on both sides, and sets out the 3CL Foundation’s position.
What the UK Is Doing
From Spring 2027, social media platforms will be legally prohibited from offering their services to anyone under sixteen. According to the government’s official factsheet published by the Department for Science, Innovation and Technology, the first set of regulations will be laid before Parliament by the end of 2026, with the changes implemented in Spring 2027. The ban applies to the major platforms that dominate young people’s online lives, with reporting from outlets including NPR naming Snapchat, TikTok, YouTube, Instagram, Facebook, and X among the in-scope services. Messaging apps such as WhatsApp and Signal are excluded, as is YouTube Kids, on the government’s reasoning that children should remain able to learn, follow the news, play, and stay in contact with people they already know.
The legislation goes beyond a simple age gate. As the government’s landmark announcement sets out, high-risk features, including livestreaming and stranger communication, will be restricted for under-sixteens across a wider range of services, including gaming sites, with stranger contact recognised as a primary route for grooming and abuse. So-called AI “romantic companion” chatbots, designed to simulate sexual relationships or roleplay, will also be required to enforce a minimum age of eighteen.
Three features define how the policy is meant to work. The first is where responsibility sits. The duty to keep under-sixteens off these services falls entirely on the platforms, not on parents or children, and families face no penalties if a child finds a way around the rules. Enforcement is the job of Ofcom, the communications regulator, which under the existing Online Safety Act framework can levy fines of up to ten per cent of a company’s global annual revenue and, in cases of outright refusal, seek court orders to block a platform in the UK altogether.
The second is age assurance. Platforms must deploy what the government calls “highly effective age assurance,” a reliable means of confirming that a user is sixteen or over. Ministers have declined to mandate a specific technology, leaving room for a range of methods, and the government has stated that this will not require a digital ID or the handing over of government-issued documents, with any data used for age checks handled under strict rules governing its use and deletion.
The third is investment in alternatives. The announcement arrived alongside substantial spending commitments set out in the government factsheet: a £500 million strategy from the Department for Culture, Media and Sport to expand enrichment activities such as sport, the arts, and youth services, a £132.5 million enrichment package from the Department for Education, and a wider programme investing over £3 billion in building or refurbishing youth centres, school sport, and grassroots facilities. This pairing of restriction with provision is significant.
A useful way to see the shape of the UK measure is through a platform that sits at its edge: Roblox. Many parents reasonably assume Roblox is social media, since children chat, form friendships, and join communities there. The UK government has confirmed it is not banned, because it is classified as a gaming platform rather than a social network. Yet under-sixteens will not be left untouched: because Britain layers function-based restrictions on top of the account ban, Roblox’s stranger-chat features are set to be switched off for minors, as are equivalent features on Fortnite and Minecraft. The Children’s Commissioner, Rachel de Souza, put the rationale plainly: boys often are not on social media at all but spend three or four hours a day gaming, on services whose features can allow “a 55-year-old in Arizona” to contact “a nine-year-old.” This is the distinguishing logic of the British approach. Rather than banning by platform category alone, it targets the risky function, stranger contact, wherever it occurs. One early win for the legislation is that it has pushed Roblox to institute new safety features after the ban was announced; surely other companies will now have to follow suit..
How It Differs from What Came Before
The UK is not legislating on a blank slate. As we examined in Online Safety: Porn and Prevention, the Online Safety Act 2023 already required platforms to protect children from harmful content and introduced robust age verification for pornographic material, applying not only to dedicated adult sites but to mainstream platforms such as Reddit and X where explicit content circulates.
The earlier framework operated on the principle of a safer presence: children could remain on platforms provided those platforms mitigated the risks they were exposed to. The new measure removes under-sixteens from the in-scope services entirely, moving from regulating what children encounter to regulating whether they may hold an account at all.
How It Differs from Elsewhere
At the European level, the picture is more contested, a tension we observed first-hand at the 3CL’s Identifying Disinformation conference, where a harmonised social media age threshold was discussed directly with MEPs in the room. The European Parliament has been weighing a harmonised digital minimum age of sixteen, with parental-consent provisions for thirteen to sixteen-year-olds. On the enforcement side, the European Commission has made a feature-ready age verification app available, an anonymous, open-source “mini wallet” that lets users prove they meet an age threshold without disclosing other personal information, and has urged Member States to roll it out by the end of 2026. Individual states are not waiting: France is advancing a ban for under-fifteens, Spain has proposed one for under-sixteens, and Greece plans to restrict under-fifteens. Notably, Estonia and Belgium have resisted the prohibition model altogether, arguing for stronger use of existing tools such as the GDPR, alongside investment in digital literacy.
Malta sits within this conversation through its green paper on social media reform, though the local context is distinct. With a very high share of the Maltese population using Facebook daily and the platform functioning as the island’s de facto public square, the costs and trade-offs of restriction land differently here than in larger markets.
Australia: The Model, and the Early Evidence
The most important point of comparison is Australia, whose Online Safety Amendment (Social Media Minimum Age) Act made it the first country in the world to bar under-sixteens from holding social media accounts, as we set out in our analysis of the Australian legislation. Canberra established the template London has now explicitly chosen to follow: platform-side responsibility, regulator-led enforcement, and penalties of up to AUD 49.5 million for failing to take reasonable steps. The Australian list captures the familiar major platforms, and notably includes YouTube, which was initially exempted on educational grounds before being added. Britain has indicated it will go further, adding the function-based restrictions on stranger contact and livestreaming described above, plus the minimum age of eighteen for intimate AI chatbots.
Because Australia’s ban commenced on 10 December 2025, well ahead of the UK, it offers the closest thing to a live experiment in the model Britain has adopted, and the early evidence is sobering. The regulator’s own three-month review found that key indicators of harm remained unchanged, with no clear decline in complaints relating to cyberbullying or image-based abuse. The eSafety Commissioner raised significant concerns about five major platforms and opened formal investigations. Crucially, the ban did not appear to be keeping substantial numbers of children off social media: in a survey of parents whose children had held accounts before the ban, nearly 70 per cent reported their children still had at least one such account. Tellingly for the UK’s gaming provisions, displaced children migrated to gaming and chat services such as Roblox, Discord, and Steam, prompting enforceable transparency notices after reports of grooming and extremist gameplay.
The picture is not uniformly negative. By mid-December 2025, some 4.7 million under-sixteen accounts had been removed or restricted, and YouGov polling found cautious optimism among Australian adults that the policy was beginning to show impact. But the headline finding, repeated across regulator data and independent research alike, is that most under-sixteens who wanted to remain on social media had done so, and the reductions in harm that justified the policy had not yet materialised. A few months is a short window, as the regulators rightly caution. The early signal, however, is that the account ban on its own does less than its proponents hoped.
The Case For
The arguments in favour are serious and deserve to be stated plainly. Early and heavy exposure to platform environments is associated with documented harms, and the algorithmic dynamics that drive engagement can pull young people toward increasingly extreme material, a pipeline we have described at length in Online Safety: Porn and Prevention. Childhood neuroplasticity makes these years a period of particular vulnerability, and the difficulty of undoing entrenched patterns once formed is well established.
A platform-side ban also corrects a long-standing imbalance. For years, parents were left to police a dozen apps child by child, negotiating with some of the most powerful companies in the world from the kitchen table. Placing the legal duty on platforms, backed by penalties large enough to matter, shifts that burden to the parties best placed to bear it. The government’s pairing of restriction with half a billion pounds of investment in youth provision also signals an understanding, at least in principle, that the question is not only what children are kept away from but what they are offered instead. Public support is substantial: the government cites nine in ten parents backing the measure.
The Case Against
The first is circumvention. As we noted in our work on the Online Safety Act, virtual private networks make it trivial to appear to be browsing from outside the jurisdiction, and VPN use rose sharply after that law’s introduction; free VPNs often carry their own dangers, logging activity, injecting malware, or redirecting users to unsafe destinations. The concern that restrictions push the most digitally literate young people into less regulated spaces is not merely theoretical. Research by youth organisation Girlguiding, surveying 2,000 boys and girls aged ten to sixteen across the UK, found that 56 per cent thought a ban would be ineffective because under-sixteens would still find a way onto platforms, and that just 15 per cent believed a ban would make them feel safer online. The Australian evidence above gives that scepticism empirical weight. The British government acknowledges this without offering much of a solution, other than that they will study this phenomenon.
The second is migration, the natural corollary of circumvention. As the 3CL argued in its response to Malta’s green paper, narrowly targeting named platforms invites a perverse compliance dynamic in which young people simply move to whatever is not on the list. The point recurred in our 3CL Conversation Series episode on Malta’s social media debate, where informal research with young people found that the platforms adults worry about most are not where socialising now happens: Facebook was dismissed as “for dinosaurs,” with Discord identified as the venue for real socialising.
A third, and subtler, concern is voice. Children’s own consultation responses have tended to ask for safer and better-designed spaces rather than exclusion from them. In the Girlguiding research, more than two-thirds (69 per cent) said they would rather know that a platform is safe than be banned from it, and 62 per cent said a ban would make them feel they were being punished. The risk in overriding it is alienation: a measure imposed on young people, framed as punishment and addressing a problem they experience very differently from the adults legislating for them, invites disengagement and quiet circumvention rather than the safer habits it intends to build. The Australian evidence shows this dynamic in motion, with non-acceptance translating directly into non-compliance. Buy-in is therefore not a courtesy but a condition of the policy working at all.
There is also a deeper point in what young people are asking for. The call for safer, better-designed spaces is more coherent than the call for exclusion from them, because online platforms are no longer a discretionary leisure activity from which a child can simply be withheld. They are, for better and worse, where much of contemporary socialisation now happens: where friendships are maintained, identities are tested, news is encountered, and the norms of public life are absorbed. A young person has to learn to operate in these environments, with all their benefits and risks, much as they must learn any other domain of adult life, and that competence is built through guided, age-appropriate exposure rather than conjured at sixteen after years of absence.
The 3CL Position
The 3CL view is that a ban can work, but only if it is carefully constructed, and that its success depends far less on the mechanics of the ban itself than on what fills the time it frees up.
The lesson running through our work, from Online Safety: Porn and Prevention to our green paper response and our reading of the Australian Act, is that prohibition is a blunt instrument that addresses where harm happens while leaving the underlying drivers untouched. A well-built ban can still buy something valuable: time, and a clearer signal to platforms that the design choices which optimise for engagement at the expense of young people’s wellbeing carry a cost. The British model improves on a pure account ban in two respects that the Australian experience shows matter. It attends to function, switching off the livestreaming and stranger contact that do concrete harm rather than relying on platform category alone, which is exactly where the Australian scope left gaps. And it places the duty squarely on the companies.
The decisive question is what happens next. A child removed from a feed is not thereby better off; they are better off only if that time is redirected toward something of substance, toward the sport, the arts, the youth provision, and the cross-generational digital education that genuine wellbeing requires. This is why the £500 million in enrichment funding may matter more than the ban it accompanies. The Australian data is, in effect, a warning about the cost of getting this wrong: where a restriction creates a vacuum and offers nothing to fill it, young people migrate to the niche sites and unregulated corners that are harder to see and harder to protect against, and the headline harms barely move. A restriction paired with real, well-funded alternatives, and with the sustained digital literacy education we have consistently argued for, stands a far better chance of changing the texture of childhood for the better.
On the evidence so far, the platforms have had their chance and, on the whole, have not taken it, and there is a defensible case for the state stepping in. But the measure of this policy will not be the elegance of its age checks or the size of its fines. It will be whether a generation of young people, given back several hours of their day, is offered something worth doing with them. That is the work that begins where the ban ends, and it is the work that will determine whether any of this succeeds.
