select search filters
briefings
roundups & rapid reactions
Fiona fox's blog
Scientists comment on the UK Government announcing a social media ban for under-16s.
Dr Madeline G. Reinecke, Postdoctoral Researcher in Collective Moral Development, University of Oxford, said:
“I welcome the government’s recognition that young people face distinct risks online, and that AI systems—in addition to social media platforms—warrant targeted regulatory attention. The new minimum age of 18 for romantic and sexually simulative AI chatbots is a reasonable step, but it addresses only a narrow slice of the risks young people themselves identify as most pressing.
“Our research, conducted with young people as co-researchers, finds that AI companionship and romance are seen by adolescents as low-likelihood concerns, while still acknowledging their potential for harm. By contrast, young people consistently flag high-likelihood, high-impact risks that the policy announced today does not address, such as over-reliance on AI for emotional support and mental health guidance, unwarranted trust in expert-sounding AI responses, and cognitive de-skilling through habitual AI use.
“We are also concerned about how age thresholds are being applied in this policy. In the United Kingdom, 16-year-olds may consent to sex and to their own medical treatment. Under this policy, however, the same young people would be prohibited from engaging in simulated romantic interaction with chatbots. There is no conclusive evidence that romantic interaction with AIs exceeds the risks young people are already trusted to navigate at this age.
“Rather than restricting access to generative AI (whether in a romantic context or wholesale), we argue that regulation should be building the conditions for competent, critical engagement: investing in AI literacy across schools and care settings, and requiring developers to design systems that support rather than undermine adolescent agency.
“In my view, bans and age gates may remove agency from young people without creating incentives for better design. Our recommendation, grounded in co-production with young people, is that the government should prioritise restrictions on the anthropomorphic and engagement-maximising design features of AI systems used by under-16s, regardless of whether those systems involve romantic functionality. This would address a far wider range of potential harms, and do so in a way that reflects the actual landscape of risk as young people experience it.”
Prof Elvira Perez Vallejos, Professor of Digital Technology for Mental Health, University of Nottingham, said:
“I don’t think a blanket ban is a good approach. This is a policy that is politically popular and stems from a well-intentioned desire to protect children, but a blanket ban is a blunt instrument that misidentifies the root problem.
“The primary threat to children online is not access to harmful content, but the data extractive business models and persuasive designs of tech platforms. By focusing entirely on access rather than platform architecture and design, the policy fails to make social media inherently safer (social media was not designed for children after all). A superior approach would be robust legislation that forces tech companies to dismantle toxic algorithms, infinite scroll, and predatory engagement loops for all users, thereby prioritising child well-being.
“The government’s policy leans heavily on a “precautionary principle,” assuming that a total ban will automatically result in a reduction of harm. However, major longitudinal studies do not support a direct, causal link between social media use itself and a mental health crisis; rather, they point to specific types of content, algorithmic targeting, and sleep deprivation. Academic bodies globally have repeatedly warned that treating all social media as a monolith ignores the vital educational, creative, and community-building benefits it provides to adolescents.
“The UK should have paused to evaluate Australia’s ongoing ban (introduced in December 2025). Emerging data from Australia’s first six months already shows severe systemic failures, proving that a rushed policy risks creating a false sense of security for parents while leaving the actual online environment un-reformed.
“Evidence suggests that outright bans have very limited efficacy and high rates of circumvention (specially with teenagers and their potential inclination for risk-taking behaviours). If the tech itself isn’t fundamentally re-engineered to be safer, children will simply find more dangerous avenues to access it (e.g., dark web). For example, driving children off moderated platforms onto unregulated, anonymous spaces where dangers like grooming and radicalisation are harder to police. It also isolates vulnerable groups, such as LGBTQ+ or neurodivergent youth, who rely on these apps for support networks, whilst creating a sudden cliff edge at sixteen when teenagers gain full access without any prior digital literacy training. Furthermore, treating youth purely as objects of protection undermines their international rights to information and digital participation.
Enforcing a ban presents major practical and privacy challenges. To verify who is under sixteen, platforms must also verify who is over sixteen, meaning adults across the United Kingdom would likely have to surrender biometric data or official identification to private tech firms. Evidence from Australia demonstrates that these bans are largely ineffective, as over sixty per cent of underage users bypassed restrictions simply because platforms failed to close existing accounts, while others easily utilised workarounds like alternative routing and shared accounts.
“The proposed path forward suggests moving away from outright prohibitions toward holding platforms legally accountable. Instead of driving children underground and creating an adult surveillance state, I would focus on monitor corporate compliance while the government legislates for safety-by-design. This would involve forcing technology firms to be more responsible and ethical by disabling addictive features like infinite scrolling and algorithmic feeds by default, ensuring a cleaner and safer online public square that prioritises young people’s well-being and mental health.”
Dr Cristina Costa, Associate Professor in the School of Education, Durham University, said:
“Blanket social media bans risk reducing what is fundamentally a systemic, technological, and commercial issue to a young people’s problem, thereby individualising the problem as if it were inherently caused by youth behaviour. The issue is multifaceted and complex and cannot be addressed by focusing on a single factor or actor alone. Many harms are embedded in platform design, including algorithmically driven interaction, weak moderation, and commercially incentivised architectures. Effective policy should therefore recognise shared responsibility across governments, technology companies, families, and users, while addressing the economic models that shape digital environments and strengthening access to digital literacy development: training for professionals and learning for young people and parents.”
Dr Liam Berriman, Associate Professor in Childhood and Youth Studies, at the University of Sussex, said:
“As someone who was involved in the government’s consultation process I’m afraid this social media ban is in danger of backfiring. I’m concerned the ban is going to make us assume we’ve fixed the problems young people are experiencing online, when actually it will just change them and make them less visible.”
“The latest research from Australia found six months into the ban 7 in 10 teenagers are still accessing social media content. Young people are likely to find a way round the new law, and when they do their profiles will be adult profiles, which means the algorithms will serve up even more inappropriate content. They’re also likely to hide their social media use from their parents and feel uncomfortable seeking help if they see disturbing things.”
“We need to instead pressure tech companies to create safe spaces, free of misinformation and harmful content. There are some aspects of the bill that could work, such as points on gaming and live streaming to stop children talking to strangers online. But that’s an example of how we’re better off targeting certain features in tech platforms rather than a blanket ban. In Australia Roblox has ended up exempt from the legislation despite being one of the most notoriously dangerous platforms for interactions with strangers.”
“The government is presenting this as a cure to the mental health crisis affecting young people, but the evidence says it’s not as simple as that. Young people’s mental health is being damaged by a raft of issues, including loneliness and exam pressure, and on the former there’s a lack of safe places for them to socialise with so many youth clubs now closed and parents uncomfortable with children playing outside independently. We need a more holistic approach and there’s a risk that taking away social media could make loneliness worse. We need to be preparing children to be digital citizens, otherwise how will they cope when suddenly at 16 they’re faced with newsfeeds that scroll infinitely and misinformation that so many adults find difficult?”
“Finding a way round VPN as the Prime Minister has suggested raises further questions – the only realistic way of tackling VPN would be if everyone had a digital ID, which in effect means giving tech companies even more of our private data. It seems unlikely voters would be happy about this.”
Dr Hisham Al-Assam, Associate Professor in Computing, University of Buckingham, said:
“The brutal truth is that “privacy preserving internet access and effective age verification” is largely a political fantasy. In practice, you can have either digital anonymity or an enforceable ban, but not both. To make such a system work, platforms would be compelled to treat every internet user as a child until proven otherwise. The result would be that millions of British adults would have to surrender passports, biometric facial scans, or financial records to technology companies simply to access online services. In effect, the privacy of the entire population would be sacrificed in an attempt to regulate the behaviour of a minority.
“The proposed solutions are also far less robust than their advocates suggest. Biometric age estimation can be deceived by high quality photographs, deepfakes, and AI generated imagery. More concerning is the creation of vast repositories of sensitive personal data. Requiring identity documents and biometric information at scale would create highly attractive targets for hackers, increasing the risk of data breaches and identity theft on an unprecedented scale.
“The deeper problem is architectural. The internet was designed to route information, not to verify human identity. Any attempt to impose state mandated identity verification across the network is therefore built on foundations that were never intended for that purpose, making such systems inherently fragile, costly, and prone to failure.
“Furthermore, policymakers appear to underestimate the technical capabilities of many young people. Large numbers of teenagers are already familiar with VPNs, proxies, and other tools that can circumvent digital restrictions within minutes. As a result, determined teenagers are likely to bypass the controls in minutes, while the burden of compliance falls disproportionately on ordinary law-abiding adults.”
Prof Matt Williams, Director of HateLab and Professor of Criminology in the School of Social Sciences, Cardiff University, said:
“The evidence base supports the view that some forms of social media use expose young people to real and sometimes serious harm, particularly through algorithmic amplification, compulsive design, harmful content, peer comparison, bullying, sexualised contact, and stranger engagement. However, the evidence does not yet support a simple claim that a total under-16 ban will, by itself, reduce harm at population level.
“Risk is unevenly distributed, harms are mediated by platform design and vulnerability, and the same platforms can also provide social connection, identity exploration, help-seeking, peer support and civic participation. Therefore, the key test is whether this particular intervention reduces exposure to harm without displacing young people into less visible, less regulated and less researchable spaces.
“Australia is the obvious comparator, but it is too early to treat it as proof of effectiveness. Its under-16 restrictions only came into effect in December 2025. Early implementation shows that platforms can remove or restrict large numbers of accounts, but it also shows the familiar problems of compliance, circumvention, platform inconsistency and incomplete evidence on actual harm reduction.
“The UK proposals around stranger engagement, gaming and live streaming are, in my view, more conceptually promising than a blunt social media ban, because they target specific risk mechanisms rather than platform labels. Much of the serious risk to children online arises not simply from being on a named app, but from particular affordances, including unsolicited contact, live interaction, algorithmic discovery, recommender systems, private messaging, sexualised chat and the collapse of adult and child publics. Restrictions on stranger communication and live interaction have analogues in existing safety-by-design systems, parental controls, default privacy settings and child account regimes, but the UK appears to be proposing something broader and more mandatory across gaming and adjacent services.
“At present, we do not have enough operational detail to assess the policy fully. We need to know which services are in scope, how “social media”, “stranger communication”, “live streaming” and “harmful functionality” will be defined, how mixed-use services will be treated, how exemptions will work, what evidence threshold Ofcom will apply, and how the government will balance age assurance with privacy, data minimisation and children’s rights. The announcement that Ofcom will conduct a rapid study on effective age assurance is therefore important, but it also reveals the central uncertainty; the enforcement architecture is being built at the same time as the policy claim is being made.
“My main concern is that a total ban may create a false sense of safety. It may reduce visible use of major platforms while pushing some young people towards VPNs, adult accounts, encrypted or fringe services, shared devices, gaming spaces, or less accountable platforms. It may also remove beneficial forms of online participation, particularly for isolated young people, LGBTQ+ young people, disabled young people, young carers, and those who use online spaces for support and identity formation. There is also a risk of widening inequalities if more advantaged families are better able to supervise, substitute and negotiate safe online use, while others are left with prohibition rather than literacy, support and accountable design.
“Age assurance is technically feasible, but it is not a panacea. Effective systems exist, including document checks, facial age estimation, reusable digital identity credentials, mobile network checks, bank-based signals, app-store level controls and layered risk-based approaches. But some children will evade them, and poorly designed systems could create privacy and exclusion risks.
“Moving forward, we must identify measurable outcomes, independent access to platform data, careful evaluation, and attention to displacement effects. We need pre-registered evaluation designs, baseline measures before implementation, audit access to recommender and age-assurance systems, and specific monitoring of vulnerable groups. We should measure not only screen time or account removal, but exposure to harm, help-seeking, wellbeing, social isolation, digital literacy, privacy impacts and migration to other services.
“The key question for me is whether this policy produces less harm, for whom, by what mechanism, and at what cost.”
Dr Astrid Van Den Bossche, Senior Lecturer in Computational Consumer Culture, King’s College London, said:
“Big Tech should absolutely not have the unfettered access they have had to children so far, but the social media ban diverts attention from the broader problems these platforms pose. Children must be given the opportunity to develop competencies and resilience in an age-appropriate manner, but a ban takes the default position that children have no right to online participation. Social media is not the same as alcohol or tobacco; many of its harms are not innate, but the result of design choices in service of extractive business models. Once children are over the age of 16, navigating social media will require more than just a health warning; we should focus on scaffolding its use appropriately.
“What is particularly worrying is that a ban does not directly invest in platforms or content provision that would be safe and enriching; it might be a clear signal to Big Tech on the direction the government wants it take, but do we really want to leave the children’s internet in their hands, again? YouTube Kids, for example, will not be subject to the ban, and, despite the additional restrictions announced by the government, the algorithms and business models that power it will likely fundamentally remain untouched.
“By speaking to and observing children playing on Roblox, I’ve learned that children’s engagements with digital spaces are far from straightforward: in their interactions (both good and bad), they develop literacies and sensibilities that are crucial to learning to navigate unregulated digital environments safely. It’s also never just about one platform; a single Roblox ‘experience’, for example, has related content circulating across various social media platforms and third-party websites. Some of it is clearly dodgy, but the children I spoke to were the first to point this out to me. This is not an argument for the status quo; it’s clear that there are real harms present that must be addressed. But I don’t see how a ban is anything but a stop-gap presented as a panacea.”
Dr Claire Haworth, Professor of Behavioural Genetics, University of Bristol, said:
“Whilst there is good evidence that some online activities are harmful for some individuals, these effects are not seen clearly at the population level. My concern is that a blanket ban may be a distraction from other important influences on young people’s mental health and wellbeing. We must be careful about over-promising what a social media ban can do to improve mental health and wellbeing.
“We must also be proactive about providing and funding alternative activities. What are we ready to offer young people that will support their social spaces and opportunities? This is especially important with the ban coming in during the winter months, when there can be fewer opportunities to socialise outside.
“The effects of a social media ban will be vastly different for teenagers already using social media who are having something taken away, compared to those younger children who will experience this new normal where social media is not available until later in adolescence. This means that there might be differences between short-term effects and longer-term effects for this younger generation. Re-imagining what a good childhood and adolescence looks like in our modern world is an exciting opportunity, and one that we should explore alongside our young people.
“In the short term we must be vigilant about unintended consequences. For some young people social media is an important source of social support and mental health support. This is likely to be especially true for particular groups. For different individuals there will be different impacts of the ban, both positive and negative. Working with our young people as individuals with different needs and interests is going to be crucial.”
Dr Sam Jones, Officer for Mental Health, RCPCH, said:
“Action has long been needed on the online harms affecting children and young people. Paediatricians are seeing growing numbers of children experiencing mental, physical and emotional harms linked to online activity, with these impacts increasingly presenting in clinics and hospitals.
“However, children’s experiences online are nuanced, and while the overall harms outweigh the benefits of children being online, reforms must reflect this complexity. The priority must be to reduce children’s exposure to harmful content online, including the impact of algorithms and addictive platform features, and create safer digital environments to better support young people’s health and development.
“We urge Government to closely monitor the ongoing impact of this ban and ensure any unintended consequences are identified and addressed, particularly for vulnerable children and young people, including those with ongoing health needs and those living in care.”
Dr Naomi Lott, Lecturer in Law, University of Reading, said:
“We know very little about the impact of a “social media ban” on young people, but we can make inferences from known benefits of other activities, and harms of excessive screen use and social media. For example, we know that young people bond better and learn better social skills if they spend time with peers in person, and we know that play and sleep are critical for children’s physical and mental health. At the same time, we know that excessive screen use has negative impacts on physical and mental health, social media exposes people to harmful content that can have significant traumatic impacts, and that social media is deliberately designed to promote its use and limit the autonomy of the user. The Academy of Medical Royal Colleges have likened screentime and social media to smoking and the need for seatbelts in cars. This is not about a moral panic, this is about knowing what is good for us and our children, and what is harmful, and taking steps to promote our wellbeing and health.
“The data coming from Australia so far is mixed and slow. What we have seen are reports that children are either able to circumvent restrictions, or have not lost the access they originally had. Due to the data that we have (or do not have), it is too early to say how effective this has been.
“I have a few concerns about a social media ban. My first concern centres on the loophole this creates for tech companies. Whilst children are on these spaces, we can call on tech companies to better regulate social media content. Following a ban, if a child circumvents restrictions and then is exposed to harmful content, tech companies may be able to hide behind the fact that these are 16+ spaces – this will place the fault at the child or their parents for accessing such content, instead of the tech companies for not governing these spaces. Another concern is how this may damage trust between children and parents. If a child accesses social media or speaks to a stranger through a gaming site, and feels as though they should raise an issue with their parents, they may feel scared of the repercussions of admitting to being on these platforms. This could limit the availability of support needed. Another concern relates to the language we are using regarding this policy – referring to it as a “social media ban” places the emphasis on children and youth and their access to this space. But it is not the children at fault, it is the tech companies. We should instead be talking about banning tech companies from gathering and using children’s data whilst promoting harmful content and habits. From what I have seen so far, I prefer the Canada proposal which keeps the responsibility on tech companies to make social media spaces safe for children and removes children from social media platforms that do not have these safety measures in place temporarily, incentivising the design and implementation of safety measures.”
Dr Oliver Davis, Reader in Biomedical Data Science at the University of Bristol, said:
“Although there have been well documented cases of individual harms, current evidence does not suggest that there is in general a large effect of social media use on young people’s mental health. This means that a blanket ban seems unlikely to have a large positive impact on a population level, and without evidence to support it, risks unintended consequences with limited potential benefit.
“At the same time, a ban for under-16s reduces pressure on social media companies to improve their platforms for everyone, and reduces the opportunity for parents and others to have open conversations with young people around how to use social media in a positive way. Those who bypass the restrictions will now be engaging unsupported on platforms that are no longer incentivised to make things safe for this age group.”
Dr Rachael Kent, Senior Lecturer and Co-Director Centre for Technology and Body, King’s College London, said:
“A social media ban for young people is an important first step, but only a first step.
“Evidence from countries like Australia that have pursued age-based restrictions suggests that bans alone are unlikely to solve the problem. Young people often find ways around age verification measures, while the underlying drivers of harm remain unchanged.
“The risk is that we frame this as a problem of access rather than a problem of platform design.
For more than a decade, researchers have documented how social media platforms are engineered to maximise engagement through personalised recommendation systems, infinite scroll, push notifications, behavioural profiling, and algorithmic feeds. These features are not neutral. They are designed to capture and retain attention.
“Restricting access may help reduce exposure to some of the documented harms associated with social media use among children and adolescents, particularly where platforms facilitate cyberbullying, social comparison, harmful content exposure, and excessive use. However, focusing solely on age restrictions risks treating the problem as one of individual access rather than platform design.
“The more fundamental issue is that many social media platforms are deliberately engineered to maximise engagement and time spent on platforms. Features such as infinite scroll, personalised recommendation systems, algorithmic feeds, push notifications, and behavioural profiling are not neutral technologies. They are designed to capture and retain attention. Young people are interacting with systems whose commercial success depends upon sustained engagement, regardless of whether that engagement is beneficial to health and wellbeing.
“My research has shown that social media platforms increasingly function as unregulated public health infrastructures. They shape how people understand health, bodies, food, exercise, illness, and wellbeing, often without the oversight we would expect of other institutions that influence public health at scale. Within these environments, algorithmic systems can rapidly amplify content that provokes strong emotional responses, rewards extreme viewpoints, or promotes unrealistic health and body ideals.
“This is particularly concerning in relation to health and wellbeing content. Young people are routinely exposed to algorithmically recommended material relating to dieting, weight loss, “clean eating”, fitness optimisation, biohacking, appearance enhancement, and wellness practices. While some of this content appears benign, research increasingly demonstrates how recommendation systems can create pathways towards more extreme material, including content that reinforces body dissatisfaction, disordered eating behaviours, exercise compulsion, and anxiety about health and appearance.
“The challenge, therefore, is not simply whether young people should have access to social media. It is whether platforms should be permitted to deploy addictive-by-design features and recommendation systems that systematically amplify potentially harmful content to young users.
“A meaningful policy response must move beyond age-based restrictions and address platform architecture itself. This includes greater scrutiny of recommender systems, stronger regulation of algorithmic amplification, transparency around engagement optimisation practices, and restrictions on design features that exploit psychological vulnerabilities in children and adolescents.
“We would not allow products aimed at young people to be designed without regard for safety standards in the physical world. Digital environments should not be treated differently simply because the harms are mediated through algorithms rather than physical products.
“If the Government is serious about protecting young people, it should draw on the extensive body of research that already exists in this area. The focus must move beyond age restrictions and towards regulating platform architecture itself. Without addressing those structural drivers, we risk focusing on who is allowed through the door while leaving untouched the mechanisms that generate harm once they are inside.”
Sharron Gunn, CEO of BCS, The Chartered Institute for IT said:
“Children and parents are dealing with enormous risks online together, driven by deliberate platform design choices like algorithmic amplification and engagement-maximising features.
“There’s strong support among our members for restricting social media use for under-16s, but they know there are challenges ahead around the technical implementation, ease of circumvention, accountability and secure age assurance.
“Government will need to work closely with technology professionals’ as it puts this legislation together, given the ambitious timescale. Above all, we’re concerned with how all technology, including AI, can be assured and trustworthy for the long-term.
“There is no such thing as a digitally native generation; new literacy skills, like identify disinformation needs to be part of education at every stage, from primary school to adult workplace.”
Dr Richard Gomer, Lecturer in Computer Science, University of Southampton, said:
What if anything would cause you concern about this policy?
“There are two broad areas of concern. The first would be the unintended side-effects of the ban itself. That would be the things like the loss of access to valuable spaces – young people do get value from using social media, often outside of the large platforms – or switching to unregulated and potentially more harmful spaces.
“The second area are the effects of the measures that are necessary to enforce the ban. Age-based restrictions will require age-gating for all users of a platform, and that process itself is potentially risky. We have seen the age verification measures introduced through the online safety act have put some people at risk. Having to prove your age by handing over things like a passport or driving license places people at more risk of things like identify theft or blackmail when those records are leaked.
There is a broader and more diffuse concern that this regulation shifts the Web farther from the original ideals of (often anonymous) free communication and connection. Many of us grew up learning from and communicating through online forums and online communities.
Is there any evidence that a total ban for under 16s could itself be harmful or have unintended consequences?
“Yes, some evidence from Australia shows that some young people will switch to other services, with less oversight, often run from other countries outside relevant legal jurisdictions.
These measures also create secondary demand for measures to circumvent blocks. Things like VPNs, or second-hand accounts, or services that undertake age-verification on their behalf. Those services may themselves be predatory and unscrupulous.
“There are risks from disrupting valuable online communities. Those may be particularly acute for minority groups such as LGBTQ+ young people, who are more diffuse and therefore rely more on the ‘long tail’ effects of the internet to connect with their peers.
“What is the role for researchers on social media at this juncture and as this policy is implemented?
We will need to follow the impacts of the changes carefully, which will mean engaging with young people to understand their experiences, but also gathering information about the structure of the communication networks that young people form between themselves. The risks of pushing communication and content to hidden or less visible spaces may reduce the salience of the issues, but will require careful research to actually check which harms remain or emerge.
Do effective age verifications exist?
“We have reasonably good ways to check if a given person is over 16 or 18 – we can rely on government-issued ID like a passport or driving license. It’s more difficult to check that the person using a service is the same person that validated their age; it can be done through e.g. facial recognition, but for long-lived accounts it could be weeks or months between doing those checks.
We’ve seen that age verification that based on things like biometrics – facial age estimation, for example – can work, but is fallible. In some cases, computer game characters have been used to verify accounts.
How feasible is it that the government can enforce this ban?
“Based on their comments, the government recognises that enforcement will not be completely effective, and many under-16s will continue to use regulated platforms, or alternatives. What’s interesting, is that the government’s explanations rely in part on “network effects” to do part of the job. Essentially, they’re saying that as young people move off these services, the value to the remaining young people will decrease. I think that’s possibly out-of-date thinking. Social media services used to be based mostly around connecting with friends; but they have shifted quite rapidly towards being algorithmic content feeds, and therefore much of the value from things like TikTok or Instagram no longer comes from having friends on the platform, it comes from interest-based content.
I think it will provide useful backup for parents, who will find it easier to push back on inappropriate use. And it will set a clear new norm.
“Complete enforcement will be completely unfeasible, but realistically it probably will have an impact on young people’s social media use.
What mechanisms and tools can they implement and how effective are they?
“The current focus has been quite narrow. Age verification is risky and fallible; and potentially quite intrusive and annoying. California has taken another approach, for example, requiring that age verification is performed at device level – on a mobile phone or computer. This is a model well worth exploring, because it could meaningfully shift access while also mitigating many of the risks associated with age verification per-service. In conjunction with mandatory content labeling, for example, a lot of effective content restrictions could be enforced on the device itself.
“Ofcom will need to identify clear scope for enforcement. It would quite reasonable for the very large platforms (and these are already defined in competition law, or the EU Digital Services Act) to be held to different standards than smaller community forums.
“We should also revisit the laws around platform liability for content. Social media companies make billions of automated editorial judgements every day. Their algorithms choose what content to show users based on what they think will generate most engagement (And hence revenue). That structurally favours content that creates outrage, often disinformation; and it often creates an appearance of consensuss where none exists, by algorithmically silo’ing people so that they see only the views that they agree with. There would be positive effects across all age groups, if platforms were held more accountable for the harmful and radicalising content that their algorithms often choose.”
Prof Sarah Bauermeister, Associate Professor, University of Oxford, said:
“Our research using data from the BrainWaves adolescent cohort study (N = 6820, Mage = 16.72) show that the associations between social media use and adolescent mental health differ by underlying motives. Avoidant-escapist use is linked to poorer mental health and wellbeing outcomes, whereas social-recreational use is associated with more favourable outcomes. Focusing on why adolescents engage with social media, rather than overall use alone, is necessary for understanding risk and informing interventions. Preprint available (https://osf.io/preprints/psyarxiv/jfs79_v1, submitted to the Journal of Child psychology and psychiatry for peer review.”
Dr Ysabel Gerrard, Senior Lecturer in Digital Communication, University of Sheffield, said:
What does the evidence say about whether this ban will be effective in reducing harm to young people?
“If the ban were to be fully adhered to by under-16s and their parents/carers, there is emerging evidence that it can help to alleviate some of the issues associated with social media, like constant contactability, endless scrolling (or ‘doomscrolling’), and the pressure to portray a certain version of yourself. Today’s youth widely report these as some of the main harms they face when using social media. However, if the ban is not adhered to and young people find workarounds – either with their parents’/carers’ support or otherwise – then these same issues will remain. Indeed, we are seeing evidence from other countries that social media bans are creating further pressures for young people as some of their peers’ parents/carers allow them to evade social media bans, while some do not, further heightening already-existing household tensions.
“I am acutely aware of the harms young people face while using social media, both because I have spent a lot of time over the past few years talking to teenagers for my research and because I am a mother myself. Social media cannot be uniformly good or bad, helpful or harmful, or positive or negative, and so it is time we stop using these binaries to try to understand it. My position against imposing a ban does not come from a denial that children face harm while using social media; rather, it comes from a longstanding frustration that a small handful of private companies have not faced meaningful regulation and are still effectively permitted to be unsafe by design. Social media was not designed with children or anyone with vulnerabilities in mind, and banning it is not – and has never been – the brave thing to do, and nor does it represent the leadership we need.
What evidence do we have from countries like Australia which have already implemented a ban?
“Emerging evidence from Australia indicates that under-16s are still on social media, but now they are using it without the necessary safeguarding measures as they will have registered themselves – either with help from their parents/carers or otherwise – as being over 16. This means they are exposed to more ‘adult’ content than they would have been if registered with their correct ages. Further, history has taught us that those who break the rules are not particularly loud about it, and so we risk learning less about what kids are doing online (and, crucially, helping them if they run into any issues) because they daren’t tell anyone that they have found a way to use social media.
What if anything would cause you concern about this policy? Is there any evidence that a total ban for under 16s could itself be harmful or have unintended consequences?
“A potential problem that is not typically reported alongside this policy is that everyone, not just young people, will likely have to provide some form of identification to social media platforms to prove that they are not under 16. It is vital that the press and government are clearer about the ramifications of this policy for everyone, and not just those under the age of 16. This realisation may change the minds of some of the celebrants.
“Further, many demonstrably harmful elements of social media – algorithmic recommendations, the endless scroll, the extent of platform companies’ data sharing practices – are things that everybody, not just those under the age of 16, requires enhanced protections from. Banning social media for a particular age group will not force powerful private companies to do anything about their unsafe design features, but this is precisely the kind of action we need from our government.
“The main problem is that a small handful of private companies have remained unchallenged for years. Unfortunately, the horse has now bolted and so we find ourselves in a situation where, rather than having consistently enforced stricter safety design features over the last few years and fought against the rising power of social media companies, it all feels too late and that a ban is the only solution.”
Dr Amrit Kaur Purba, Assistant Professor, Wellcome Fellow, Academic Lead of the Digital Determinants of Health Hub at the London School of Hygiene & Tropical Medicine (LSHTM) said:
“There is a strong public health rationale for taking action to protect children online. We know that many young people are exposed to harmful content online, and growing evidence suggests that digital environments can influence a wide range of health, behavioural and developmental outcomes. Many parents will therefore welcome efforts to strengthen protections for children. However, it is important that we do not view social media restrictions as a simple solution to a complex problem.
“From a public health perspective, social media restrictions should be understood as complex systems interventions. Digital platforms are embedded within wider social, technological, commercial and political systems involving young people, families, schools, governments, advertisers, content creators and technology companies. When interventions are introduced into complex systems, the system adapts. Young people may migrate to alternative platforms, gaming environments, encrypted services or AI-based tools. Companies may redesign products, alter recommendation systems, strengthen age-verification processes, or shift activity into less regulated spaces. As a result, harms may be reduced, but they may also shift, evolve or be redistributed in ways that are difficult to predict.
“Importantly, success should not be judged solely by changes in mental health or reductions in screen time. Social media is increasingly recognised as a broader determinant of health. Digital environments can influence exposure to violence, self-harm and harmful content, substance use, sleep, physical activity, educational engagement, social relationships, social connectedness, identity development, civic participation, inequalities, and young people’s wider opportunities to thrive. The key question is not simply whether social media use falls, but whether children’s lives become healthier, safer and more equitable as a result.
“The other crucial point is that a ban should be viewed as one step within a broader public health strategy, not the final destination. Many concerns about social media arise not simply because platforms exist, but because of how they are designed. Features such as algorithmic amplification, infinite scrolling, engagement-driven recommendation systems, livestreaming, stranger contact, persuasive design and data-driven business models can shape what young people see, how long they remain engaged, and the risks they encounter. If we focus only on restricting access, we risk addressing symptoms while leaving some of the underlying drivers of harm untouched.
“The government’s decision to combine age restrictions with measures targeting specific platform features is therefore particularly important. The key question is not simply whether a ban is introduced, but how effectively it is implemented, enforced and evaluated over time. Ultimately, the success of the policy will depend not only on whether it reduces platform use, but whether it reduces harmful exposures, improves health and wellbeing, avoids widening inequalities, and responds effectively to how young people and industry adapt over time. The priority now should be rigorous independent evaluation so we understand not only whether the policy works, but how it works, for whom it works, and whether there are unintended consequences across the wider digital ecosystem.
“Ultimately, the goal should not simply be to create children who spend less time online. It should be to create digital environments that are healthier, safer and fairer for children. A ban may be an important step towards that goal, but lasting progress will also require platforms to become safer by design.”
Dr Siamak Shahandashti, Senior Lecturer in Cyber Security & Privacy, University of York, said:
“There has been some recent research published on the effectiveness of deployed age verification mechanisms in adult content websites by researchers in Politecnico di Milano. The paper can be found here: https://arxiv.org/pdf/2606.08667
“The results in this paper are some of the first empirical evidence we have for the effectiveness of these deployed methods, which include age estimation (e.g. via selfie camera), ID verification (e.g. through physical ID), email-based (e.g. through age of email account), and credit card based (e.g. through credit checks).
“The authors found low or medium robustness (meaning either more than 75% of providers or 25-75% of age verification providers can be bypassed, respectively) for almost all methods except credit checks. Crucially, the tools and expertise required to bypass these methods were judged to be within reach of “motivated minors”.
“The authors conclude that “regulation-mandated age verification for adult websites currently functions as compliance theatre” (i.e. merely a tick-boxing exercise without proven effectiveness).
“In my view, although this research paints a dim picture of the state of the art in deployed age verification and demonstrates that most current age verification solutions lack robustness, it also provides a glimmer of hope and shows that achieving sufficiently robust age verification is not technically infeasible, and if clear targets for the required level of effectiveness are established, some of these systems (especially those who tie digital identity to established physical identity documents) could be refined to meet acceptable standards in near future.”
Prof Miranda Pallan, Professor of Child and Adolescent Public Health, University of Birmingham, said:
“There is mounting evidence for the harms of social media in young people but this also extends to vulnerable adults. The embedded and ubiquitous use of social media within our society, along with the vested interests of large tech companies, makes this a complex issue.
“The measures that the government are introducing will limit ‘end user’ access, and the additional plans for blocking livestreaming and stranger communication will start to address these specific online safeguarding issues. The government is also attempting to prevent a cliff-edge at 16 by blocking livestreaming and stranger engagement by default until the age of 18.
“As with other major public health concerns, the measures that are being introduced are likely to contribute to addressing the issue of social media harms in young people, but still need to be embedded within a wider approach to social media within our society. Regulating tech companies to stop and remove harmful online content is still imperative, as this will make online and social media spaces safer for everyone, and so this approach needs to go hand in hand with the measures being introduced.
“The experience of the Australian ban also needs to be considered. It is evident that, while social media access has reduced for under 16s in Australia, many children and young people are still able to access various platforms. Thus, with the introduction of these measures in the UK, the government needs to be mindful that those who are most vulnerable to the harms of social media may not be protected by these measures and ensure that there are mechanisms in place to support these young people.”
Dr Lizzy Winstone, Senior Research Associate at the University of Bristol Medical School and NIHR School for Public Health Research Postdoctoral Launching Fellow, said:
“My main concern is that a ban could create a false sense of safety. The digital environment will still exist when young people turn 16, and risks will not disappear simply because access is delayed. There is risk that young people may enter these spaces later but with less experience, less digital literacy, and less adult guidance. This could undermine resilience if the policy reduces opportunities for young people to practise safe, supported online participation. Increased digital literacy is an essential complement to industry regulation and safer design of platforms (https://doi.org/10.2139/ssrn.6406621; https://www.bristol.ac.uk/policybristol/policy-briefings/algorithmic-literacy-wellbeing/).
“The evidence does not yet allow us to say confidently that a blanket under-16 social media ban will reduce harm to young people. There is good evidence that some online experiences are harmful. Exposure to distressing content, unsolicited contact from strangers, pressure to be constantly available, fear of negative evaluation, privacy risks, and time displacement can all contribute to digital stress (https://doi.org/10.1177/02724316221105560).
“We have evidence that cyberbullying is associated with poorer mental health and self-harm, and that most perpetrators of cyberbullying are also victims themselves (https://doi.org/10.1016/j.jadohealth.2024.04.004). The evidence also shows that social media use is not one single behaviour (https://doi.org/10.1002/jcv2.12071). Young people use platforms for friendship, identity development, social and emotional support, entertainment, information and connection, despite the risks (https://doi.org/10.1177/02724316221105560; https://doi.org/10.31234/osf.io/ye869_v2; https://doi.org/10.2139/ssrn.6406621). A total ban may reduce some exposures, but it may also displace young people to less visible, less regulated spaces to meet their needs.
“It is too early for evidence from Australia to confirm whether bans improve mental health or safety. Australia has implemented age restrictions, but what we mainly have so far is evidence about enforcement, compliance and circumvention. We do not have evidence of reduced anxiety, depression, self-harm, sleep problems, grooming, or exposure to harmful content.
“The proposed measures on stranger engagement in gaming and live-streaming platforms are potentially more targeted and, in principle, more consistent with the evidence. Our research with 13-14-year-olds highlighted unsolicited contact from strangers and expectations of availability through private communication as specific sources of digital stress (https://doi.org/10.1177/02724316221105560). Restrictions on adult stranger contact, location-sharing and disappearing messages are welcome, provided they are carefully designed and independently evaluated.
“Different young people use social media differently and feel the benefits and risks unevenly. It will be important to see how the policy will apply to and impact children with additional needs, young carers, LGBTQ+ young people, disabled young people, and those who rely on online communities for support.
“Researchers will need to evaluate implementation, listen to young people’s experiences, examine inequalities, and track both benefits and unintended harms over time.
“We also need to look at the system as a whole, e.g., if social media is removed or restricted, what is being provided in its place? Young people need safe spaces for social connection, creativity, play, information, participation, and support. Real, equitable investment in offline activities is critical.”
Dr Thomas Lancaster, Principal Teaching Fellow, Department of Computing – Faculty of Engineering, Imperial College London, said:
“I do understand why parents want decisive action. What I’m not clear on yet is whether this is the right action. Australia’s restrictions on social media for under 16s are recent and there’s little evidence yet that young people are safer and happier as a result. What the Australia ban has shown is that technology platforms can ban accounts at scale if they are motivated or legally required to do so, and they do already have the infrastructure to quickly restrict access to UK children. At the same time, the ban will never be 100% complete, and we don’t know how many children have workarounds. At this stage, I’d consider a similar ban in the UK as being as much experimental as it is based around evidence.
“A complete ban is a blunt instrument. It’s really specific features on the social media platforms that are of concern. A child livestreaming to an unrestricted audience or sharing their location with unknown adults opens them up to predatory behaviour. A gamified social media platform, where children are rewarded for continual posts and interaction with others, does encourage them to let social media take over their lives. That’s dangerous for people of all ages. But those are individual features that need to be reconsidered. It doesn’t mean that the social media platform itself is inherently bad.
“There really are welfare concerns as well. If we push away young people from peer support, maintaining friendship, even being able to connect with like-minded people to provide help with their disabilities, it doesn’t mean that they’ll instead seek help from adults. They might end up being socially isolated or moving to riskier platforms. Without careful consideration beyond the bluntness of a ban, we run the risk of children moving from one challenging situation to another. Without firm evidence that bans work, it is the unintended consequences that are the most worrying.
“Regarding research and data, I encourage the Government to put a plan in place now to evaluate the success of the ban, rather than, as so often happens, rather than relying on anecdotes or trying to reverse engineer it after declaring the policy is in place. From a research agenda, this means we need baseline measures now, along with transparent and independent access to data. I’m hopeful that the ban will lead to harm reduction overall, but there is a trade-off with benefits lost, and both sides of the discussion need to be equally tracked.”
Dr Harry Dyer, Associate Professor in Education at the School of Education and Lifelong Learning, UEA, said:
“The announcement from Keir Starmer today regarding a social media ban for young people is a worrying step that will not only fail to address the risks young people face online, but will likely put them at further risk.
“It is also failing to protect young people from online harms and predatory social media companies which will be let off the hook for how they deal with the young people, who will continue to use these platforms regardless of a ban.
“This policy sounds like action, while guaranteeing inaction on the things young people in our research have told us actually matter to them. We absolutely need stronger protections for young people online.
“But a ban is the wrong tool for the job, and it doesn’t do a good job of holding platforms accountable. The announcement of a ban for young people is essentially setting up a bouncer for a building that’s already on fire. Checking ID on the door isn’t going to put the fire out, and it’s going to do nothing to protect the people already inside.”
“In my first book, Designing the Social (2020), I explored at length just how difficult it is to define social media meaningfully, and how varied young people’s engagement with platforms actually is.
“Children use collaborative Google Docs as social spaces, Minecraft servers, WhatsApp groups, Roblox chat, and fanfiction communities, and a range of other digital spaces to act and interact with others.
“Today’s ban targets a handful of platforms without grappling with the ways social media are woven into children’s lives in complex ways beyond the ‘big’ platforms this ban addresses. Meanwhile young people continue to use digital spaces for a range of reasons, and need protections built in to help them navigate these spaces in ways that respect their rights and needs.
“We also know from emerging data coming out of Australia, who have enacted a similar social media ban for young people, that their ban has not effectively stopped young people using social media.
“Instead, young people are turning to VPNs, using fake accounts, fooling photo verification, and engaging in riskier practices to access these platforms. Data from Australia suggest as many as two thirds of young people are still using social media despite the ban, and that’s likely to continue to be the case here in the UK, suggesting that this won’t actually be effective in the task it sets out to accomplish.
“Meanwhile, the platforms are less incentivised to actually protect young people, who will still be there regardless. Rather than being required to build in safety features and being correctly regulated for the ways they process and store young people’s data, this announcement of a ban is letting them off the hook and responsibilising young people for how they use digital spaces.
“The ban doesn’t take into account the need to address the rights of young people. what rights do children have to privacy, to education, to participation in digital life, to have their data handled responsibly? A response is needed that starts with a consideration of what the rights of the child are, and what we can do to better ensure platforms don’t violate these rights.
“We should be placing the responsibility on platforms to design safer systems for the young people who absolutely will continue trying to access them, even with a ban in place.
“Many of the features needed to protect the young people who are on these platforms will likely benefit all users, including clearer terms and conditions, better processing of data, restrictions of algorithmic recommendations, rights to privacy and removal, and legal standings to regulate and challenge platforms. Rather than regulating these companies to build in features that will benefit all of us, including young people, today’s announcement has let the platforms off the hook.
“This will likely also create implications for how easily young people feel they can talk to adults about what is happening online. The ban creates a legal minefield for schools, who will likely struggle to teach digital literacy in a meaningful way.
“Our recently completed UKRI-funded grant, exploring young people’s perspectives on AI in education, highlighted how education already struggles to addresses digital experiences of young people in inadequate ways.
“Current approaches towards talking about social media in schools are often detached from the actual experiences and needs of young people, and young people are left with a lot of questions.
“We found that young people really wanted to talk about the role of social media and technology in their lives, but had little avenues to do so.
“Young people will continue to be curious about social media, and this ban does nothing to help them navigate this curiosity.
“We know that abstinence approaches in sex education do little to protect curious young people, and can actually create a more harmful environment. This ban takes a similar approach, restricting young people without given them the educational tools to understand digital environments safely.
“This announcement of a ban will likely not help schools to address the needs of young people, who will have a lot of questions, but now will have less certainty about who they can turn to if something goes wrong.
“Schools and teachers will likely be uncertain about what they can actually say, so it’s also likely to cut off vital avenues for discussions. Teachers won’t know what to say beyond language of the ban, and kids will feel they can’t really talk to teachers. It’s essentially cut off a vital lifeline for kids in the name of ‘safety’.
“What is needed are legal obligations on companies to build children’s safety in from the ground up; serious investment in digital literacy education that meets young people where they actually are; and proper resources for parents who are currently left to navigate this alone.
“Stronger and clearer enforcement of the Online Safety Act, mandatory safety-by-design requirements for platforms, and a properly funded digital literacy curriculum would be a meaningful start, and more effective than a ban.
“Instead, we have landed on an abstinence approach that will not stop anything, and that leaves young people more exposed, less supported, and with fewer adults they can trust.”
James Stevenson, Technology-Facilitated CSEA Data Specialist, Childlight Global Child Safety Institute, hosted by University of Edinburgh, said:
“Childlight welcomes today’s news on social media restrictions as an important recognition of the scale of harm many children face online. While no single measure will solve online child sexual exploitation and abuse, our research shows children experience these harms at a rate equivalent to around 10 victims every second globally, and the status quo is failing too many children.
“We will want to see the detail of the proposals, but the evidence is clear that stronger action is needed from governments, regulators and technology companies. We increasingly understand online child sexual exploitation and abuse as a public health issue, with serious consequences for children’s mental health and wellbeing, including trauma, anxiety and self-harm. If these measures succeed in reducing children’s exposure to online exploitation and abuse, they have the potential to improve health outcomes and spare many young people from harms that can affect them throughout their lives.”
Prof Dawn Watling, Professor of Psychology, Royal Holloway, University of London, said:
“The Prime Ministers ban for under 16s is not the solution.
“Children already bypass the age restrictions of social media sites and in some of our own research, 39% of seven to 12 years olds accessed social media, with 31% reporting they owned at least one social media account. This is supported by Ofcom reports.
“The government’s ban is in no way grounded in evidence. Research evidence supports that when children’s mental health and wellbeing is poor, they use social media more.
“However, evidence that more frequent use of social media, not just time online, but engaging with online content and posting, is weak and often non-existent as having a contributing role to declining mental health and wellbeing.
“Therefore, it’s important to focus on children and young people’s changing online behaviours to support them.
“The government needs to focus efforts to hold social media companies accountable for not allowing inappropriate and harmful content to be displayed; they need to change their oversight and algorithms. The government should have waited for more evidence.
“Researchers need to continue to assess adolescents access and engagement with social media and assess links with mental health over time and other factors that may may be impacted, such as sleep patterns and attention.
“We want to explore what it means for adolescents and the impact on additional pressures and opportunities when they turn 16, such as being able to engage with social media, have the right to vote, and gain a driving licence.”
*UPDATED COMMENT* Rafe Clayton, Senior Lecturer in Media Practice at the University of Leeds, said:
“Today’s announcement marks a landmark policy change, that many parents will welcome. For years now, the dominant public message from big tech and Government has been that the digital world is beneficial and productive. Whilst true to an extent, the public have often expressed serious concerns around negative screen time impacts, as shown by our 2022 University of Leeds study.
“A cultural response is now occurring, as young people and parents are recognising the harms that the digital world brings. It is right that the Government have recognised this and are acting upon the worries of the population. In my view, the global evidence base supports these restrictions.”
Prof Amy Orben, Research Professor and Programme Leader at the MRC Cognition and Brain Sciences Unit, University of Cambridge, said:
“The UK government has just announced a social media ban for under 16s. Whether this is a good or bad policy decision depends on what we consider to be the ultimate goal of this ban.
“On the one hand, this ban will not solve our collective concerns about the increasingly digital childhoods experienced in the UK today. We know from the Australian ban that current enforcements are incomplete, and the majority of young people are still online at similar rates. Evidence synthesis from my team and others shows that we should likely not expect substantial boosts to well-being or mental health in the short term, or large changes in behaviours or rates of parental conflict.
“However, a ban is likely to change public perceptions, and make social media use less acceptable in younger age groups. This is an important first step in public health education and behavioural change. It can also minimise instances of individual harms for young people who cease engaging with platforms, and over time it can, if done right, change our culture around social media use among certain age groups.
“First and foremost, the ban is a recognition by government that previous policies to make social media safe have not worked as planned. Banning something for those most vulnerable is a good step if it cannot be made safe. But we know why social media is at times unsafe for not just children but adults as well: this includes harmful content, conduct or communications, as well as design features that make it harder for us to disengage even when we want to. We have failed to adequately address these.
“By banning under 16s, we are constraining their rights to join in with this part of our digital society, because we have lost trust that we can make social media safer for us all at any acceptable rate. I feel a deep sense of disappointment about this.
“I welcome the government’s further announcement that it will introduce restrictions on specific design features across the online world that are particularly high risk, such as AI chatbots or engaging with strangers online. However, again, the question will be about the effectiveness of its implementation.
“Independent evaluation of whether this ban works will be crucial. The UK government needs to stand by its word and show flexibility in addressing digital harms, as it will be an ongoing process – not a one-stop solution.”
Prof Pete Etchells, Professor of Psychology and Science Communication, Bath Spa University:
“While there is general agreement within the scientific community that more needs to be done to ensure that digital platforms are safe by design for children, we still don’t have much solid, objective evidence on the impacts – either way – of age-based social media restrictions. That doesn’t mean that nothing should be done; instead, given that uncertainty, it is important to move forward in a careful, deliberate way. As we learn more about the specifics of how this ban will be implemented and work in practice, it will be important to keep a close eye on the evidence base that is developing around it, so that the potential unintended consequences that many worry about can be identified and reduced early on. Complex issues require complex solutions, so I think it’s important that the ban is seen as one part of a broader set of efforts to improve young people’s digital lives. This also means putting a greater emphasis on building digital and media literacy skills, so that by the time children are old enough to access these platforms, they are well prepared to use them.”
Prof Victoria Goodyear, Professor of Physical Activity, Health, and Wellbeing, University of Birmingham, said:
“Social media is currently embedded in young people’s daily lives, and evidence highlights both benefits and risks. It can support connection, learning and access to information, but there is also consistent evidence linking high and unstructured use with poorer sleep, reduced physical activity, and distractions from learning—all of which can negatively affect wellbeing. Much of the current research focuses on overall screen time, and there is broad consensus that limiting excessive use remains important.
Current UK guidance reflects this balance. The CMO guidance emphasises setting clear boundaries around screen use, prioritising sleep, and supporting children to develop healthy habits. At the same time, the Online Safety Act places increasing responsibility on technology companies to design safer platforms and reduce harmful content. The forthcoming screentime guidance by the Department for Education for 5-16s will be increasingly important to help parents make informed decisions about supporting their children’s healthy screen use.
“While school or national bans on smartphones and social media are being introduced in some contexts, evidence suggests these measures alone are unlikely to be sufficient to impact on young people’s mental health and wellbeing, and associated education and health outcomes (e.g., attainment, physical activity, sleep). In a digital society, longer-term approaches are needed. These should combine safety-by-design—reducing persuasive and attention-grabbing features—with education that builds young people’s digital skills, confidence and autonomy, particularly as they transition into a technology filled world from age 16.
“Crucially, policy and practice should be supported by ongoing evaluation to strengthen the evidence base and ensure that interventions are both effective and proportionate.”
Prof Andy Miah, Chair of Science Communication & Future Media, University of Salford, said:
“Banning social media for under 16s is the worst possible reaction to concerns about harmful and unhealthy habits online. We’ve spent 20 years ignoring the risk; schools and parents haven’t known what to do, and this is a policy born out of desperation arising from the failure to be bold in guiding young people towards healthier and empowering habits during their time at school.
“We absolutely need more evidence to understand the impact of a ban but, in the absence of clear evidence, we need proactive interventions that support healthy online behaviour. What happens when a child turns 16? Are they just turned out into the Wild West of the internet and expected to protect themselves? I fear this ban is simply kicking the risk down the road and I’ve heard nothing about positively empowering young people to become more resilient to the risks. This should be the policy focus.
“I don’t think we know enough about the impact of the policy, but we can be sure that the consequence is both impractical to enforce and stifling of open conversation. Habits of using digital platforms will go underground and, as such, we’ll know less about how children are using social media.
“Also, the policy doesn’t get to grips with what social media is and this is a fundamental problem. For example, we know that young people are using AI like social media now and there’s even less known about that.
“Obvious examples of risk arising from a ban include the young people who, currently, have their social network through their digital worlds. Not everyone has a supportive friendship group in person and, while that may seem sad, removing social media access for such people means social isolation. So, we’ll likely see a number of mental health concerns arise as a result of this ban.
“Perhaps the biggest loss from this ban is the positive conversation we could be having about the remarkable technology that everyone has in their hands now. These powerful devices could be used to change the world and we’ve not had that conversation ever with children. It’s just been completely neglected. “
Prof Andy Phippen, Professor of IT Ethics and Digital Rights, Bournemouth University, said:
“On effectiveness, I would be cautious. The evidence does not suggest that a blanket ban, on its own, will substantially reduce harm to young people. It may reduce some forms of exposure on the largest platforms, but harm is not produced simply by access. It is produced by design features, recommendation systems, commercial incentives, weak moderation, poor reporting routes, and the absence of trusted support when things go wrong.
“Australia is the obvious comparison, but it is still too early to draw strong conclusions. Its ban only recently came into force, so we do not yet have robust longitudinal evidence on whether it reduces harm. What we do have is evidence of implementation difficulty: questions about age assurance, circumvention, VPN use, false accounts, and whether young people simply move to less visible or less well-regulated spaces. That should make us careful about presenting Australia as proof of concept, or whether a “stronger” approach to the Australian approach will be any more successful.
“The proposed restrictions around stranger engagement on gaming and livestreaming platforms are potentially more targeted and, in my view, may be more promising than a blanket social media ban. Limiting unsolicited contact, disabling location sharing, restricting disappearing messages, and reducing adult access to children in high-risk spaces are all more closely connected to specific patterns of harm. Similar ideas already exist in online safety regulation and platform safety design, but consistent implementation across gaming, streaming and messaging environments remains patchy.
“At this stage, I do not think we have enough detail to fully assess the policy. We need to know which services are in scope, how “social media” is being defined, what age assurance will be required, what data will be collected, how appeals will work, how Ofcom will enforce compliance, and how unintended consequences will be monitored. Without that detail, it is very difficult to separate a serious safeguarding intervention from a politically attractive announcement.
“My main concern is that a total ban risks becoming a form of policy theatre. It may reassure adults while shifting attention away from the harder questions about platform design, regulation, education, youth support and enforcement. There is also a real risk that some young people will be pushed into more hidden spaces, become less likely to disclose harmful experiences, or lose access to peer support, identity exploration and information. For vulnerable young people in particular, the internet is not only a place of risk; it can also be a place of connection and help.
“The role of researchers now is not simply to line up for or against the ban. We need to ask better questions: what harms are reduced, for whom, in what contexts, and at what cost? We also need to track displacement effects, children’s lived experiences, parental responses, platform compliance, and whether the policy actually changes the systems that produce harm. Crucially, young people themselves need to be part of that research, not just the object of it.
“On age verification, effective tools exist for some contexts, but there is no perfect, universal solution. Document checks, facial age estimation, mobile network checks, payment-card checks and trusted digital identity systems can all play a role, but each comes with trade-offs around accuracy, privacy, inclusion, cost and circumvention. The more robust the system, the more intrusive it may become; the less intrusive it is, the easier it may be to bypass.
“So yes, government can make access harder. It can require platforms to adopt stronger age-assurance systems, audit compliance, impose fines, and restrict high-risk design features. But “harder” is not the same as “impossible”. Enforcement will depend on platform cooperation, Ofcom’s capacity, technical standards, and whether the policy can adapt as young people and platforms change behaviour. A ban may be enforceable enough to alter the mainstream market, but it should not be sold as a clean technological fix.”
Prof David Ellis, Chair of Behavioural Science, University of Bath, said:
“This ban is based on worry, not evidence. The evidence base as it stands suggests social media has a minuscule effect, if any, on teenagers—particularly once you account for the other factors we know shape childhood development.
“It’s also unlikely to be straightforward to enforce, given what we’ve seen elsewhere, and it risks pushing teenagers towards less regulated parts of the internet. Worse, it lets social media companies off the hook: they can divert resources away from making platforms safer, despite the fact that many young people will simply remain on them.
“This is what happens when politics is put before evidence-based policy. Rather than tackling the difficult question (how to make the online world safer), you sledge hammer yourself into a worse position than when you started.”
(original comment sent on Sunday) Prof Alan Woodword, Professor of Computer Science, University of Surrey, said:
“What is conspicuous by its absence is how we achieve this ban. An outright ban rather than policing the product safety risks mandating something that will fail, and thus not actually achieve the objective which has to be keep the children safe.
“The evidence from Australia suggests that bans are not effective, so it must be better to address the problem with an approach that will work, and will protect children.”
Dr Catherine Sebastian, Head of Evidence for Mental Health at Wellcome, said:
“This is a once-in-a-generation opportunity to gain a better understanding of youth mental health. We do not know how this ban will impact teenagers’ mental health. It is crucial that scientists closely observe what happens to inform future policy.
“Wellcome will be funding nationwide studies to independently monitor the impact of the ban. Researchers will examine whether there is an improvement in teenagers’ mental health – and if so, why. Is it, for example, that teenagers are socialising in real life more, or is it that they are getting better sleep? Young people want to know the answers to these questions, too, so we will work with them to design the studies.
“In addition, this evaluation will help us suggest new ways to support young people. As countries around the world explore the same path, the UK can lead the way in providing rigorous evidence on how the digital world affects mental health.”
Expert comments from Sunday (when there was media speculation of the ban) can be read here:
expert reaction to speculation of a ban on social media for under-16s
Declared interests
Dr Madeline Reinecke: Madeline G. Reinecke was the Principal Investigator of an AI Security Institute Challenge Fund Grant (“Co-Designing Adolescent Safeguards for Generative AI: Mapping Risk and Prototyping Mitigations with Young People”). She interned at Google DeepMind in 2022.
Dr Hisham Al-Assam: I have no interests to declare
Dr Liam Berriman: No conflicts of interest
Prof Matt Williams: I am on the Ofcom (Wales) Advisory Board, and am a founder and director of nisien.ai, but this comment is in my capacity at Cardiff University.
Dr Astrid Van Den Bossche: No conflicts of interest to report.
Dr Claire Haworth: I have no conflicts of interest to declare.
Dr Sam Jones: Working party for CYP APEX (the training course for CYP in mental health crisis) but no financial payment (all volunteer work)
Dr Naomi Lott: No conflicts to report.
Dr Oliver Davis: None
Dr Rachael Kent: Alongside her academic work, she is the founder of Dr Digital Health, through which she advises businesses, the NHS, government, and policymakers on the harms of digital technology and strategies to support healthier digital habits. Dr Digital Health work is all her own research, and none is commercially vested or driven.
Dr Richard Gomer: I am a district councillor and member of the Liberal Democrats; I do not speak for the party (many of these points directly contradict our spokespeople!)
I am the Mayor of Eastleigh 2026-27. It isa civic mayoralty, so in that role I am non-partisan.
I am a non-executive director of Defend Digital Me, an organisation that campaigns for digital rights.
Between 2012 and 2018 I worked with Facebook on their “Trust, Transparency and Control” labs as an external advisor. I do not currently have any research funding or other income from Meta.
Prof Sarah Bauermeister: No declarations of interest, I am funded by BrainWaves and Dementias Platform UK.
Dr Ysabel Gerrard: Alongside my role as a Senior Lecturer in Digital Communication at the University of Sheffield, I am an unpaid member of Meta’s Suicide and Self Injury (SSI) Advisory Board.
Dr Amrit Kaur Purba: AKP is funded by the Wellcome Trust and leads the Digital Determinants of Health Hub at the London School of Hygiene & Tropical Medicine. She has provided unpaid advisory input to the UK government and regulatory bodies, including the Home Office, the Department for Science, Innovation and Technology, Ofcom, the Metropolitan Police, and No 10 Downing Street, on issues related to online harms, digital regulation, youth violence, and adolescent health. She also holds unpaid advisory roles with the Alliance 4 Children and the UK Department for Education’s Technical Advisory Board (Education and Outcomes Panel) and has contributed to advisory activities related to youth digital technology use and public health through WHO and United Nations initiatives. She has received honoraria from academic institutions, government bodies, and public health organisations for invited talks relating to social media, online harms, and adolescent health.
Dr Siamak Shahandashti: My research is currently supported by the UKRI and RAI UK/EPSRC, the usual funders of physical sciences in the UK, however the projects I am currently involved in is on trustworthy AI (e.g. PHAWM, see website for more info) are not related to age verification, so I do not see any relevant interest to declare.
I was part of the UK Information Commissioner’s Office Privacy-Enhancing Technologies Expert Forum (2021–2022), but I don’t see that relevant here either.
Prof Miranda Pallan: I do not have any DOI.
Dr Lizzy Winstone: None
Dr Thomas Lancaster: No COIs
Dr Harry Dyer: I’ve not been funded by any social media companies to do research, my research has been funded by the UK government.
James Stevenson: None
Prof Dawn Watling: No declarations
Prof Amy Orben: “Co-principal investigator of the Wellcome funded social media reduction trial in Bradford; member of the Austraian eSafety Commissioner’s Social Media Minimum Age Evaluation Academic Advisory Group; Director of the DSIT research commission “Feasibility Study of Methods and Data to Understand the Impact of Smartphones and Social Media on Children and Young People”; Member of DfE Science Advisory Council and DSIT/DCMS College of Experts; ESRC Smart Data Research UK Programme Board member, Digital Futures for Children Advisory Board member; In 2023 I gave paid talks to SWGfL and Apple University; I have received funding or consultancy payments from UKRI, Wellcome Trust, Jacobs Foundation, Huo Family Foundation, UK Department of Innovation Science and Technology, Prudence Trust, National Institute of Health, Emmanuel College, University of Cambridge and Barnardo’s.”
Prof Pete Etchells: Member of the DSIT Expert Panel for Growing Up in the Online World; Member of the Australian eSafety Commissioner’s Social Media Minimum Age Evaluation Academic Advisory Group; Member of the Department for Education’s 5–16 Screen Use Expert Advisory Group; Member of the DCMS College of Experts; Author of Unlocked: The Real Science of Screen Time (and how to spend it better); Author of Lost in a Good Game: Why We Play Games and What They Can Do For Us
Prof Andy Miah: None
Prof Victoria Goodyear: Related work has also been funded by the NIHR, ESRC, Birmingham Alumni Funding, and Research England, and VG has been part of a consortium for the Department for Science Innovation and Technology and provides advice to Department for Education on adolescents. VG currently receives funding from Department for Education for a systematic review on digital media. All funding is paid to The University of Birmingham.
Prof Andy Phippen: I’m a trustee of SWGfL, who are an online safety charity.
Prof David Ellis: I receive funding from UKRI and was a member of a government-commissioned research project led by Cambridge – more details here https://www.bath.ac.uk/announcements/university-of-bath-research-to-examine-impact-of-smartphones-and-social-media-on-young-people/
Professor Alan Woodward: No conflicts to declare
Rafe Clayton is currently working on a project funded by the 1001 Critical Days Foundation about the impacts of screen time on babies under the age of two. The study referenced was funded by Research England’s Policy Support Fund.