Introduction

Recently, there has been significant public discourse about the Australian social media ban for under-sixteens and calls for similar measures elsewhere in the world. This has culminated in the announcement of plans for a similar ban in the UK (Department for Science, Innovation and Technology, 2026).

The idea of such a ban is emotionally appealing and rhetorically convenient, especially as advocated by parents affected by high-profile tragedies linked to social media. While their grief is undoubtedly genuine, it does not guarantee the effectiveness or practicality of the solutions proposed. This essay will demonstrate that a social media ban for under-sixteens is largely a product of emotional and moral narratives, lacks strong supportive evidence, and is unlikely to meaningfully reduce harm. Instead, it risks distracting from the need for genuine solutions and could permit both existing and new harms to persist.

• 1. A Policy Built on Nostalgia, not Evidence
• 2. The Political Framing: “Giving Childhoods Back”
  • 2.1 Grief, urgency, and stacked correlations
• 3. Misdiagnosing the Problem: Harmful Content Is Not Inevitable
  • 3.1 The ban assumes teenagers are uniquely vulnerable
  • 3.2 The evidence for adult harm is at least as strong
  • 3.3 Young people do not believe it will be effective and do not want it
• 4. Practical Failure: Age Verification Does Not Work
  • 4.1 Age verification is unreliable
  • 4.2 Age verification creates new risks
• 5. New Harms Created by the Ban
  • 5.1 Isolation of vulnerable young people
  • 5.2 Misleading product‑safety analogies (“take it off the road”)
  • 5.3 Social media is not a separate sphere, but a discrete channel for existing relationships
• 6. A Better Alternative: Literacy, Regulation, and Shared Responsibility
  • 6.1 Platform regulation and accountability
  • 6.2 Digital literacy training measures for teenagers and adults
• 7. Conclusion: Symbolic Politics Will Not Protect Children

1. A Policy Built on Nostalgia, not Evidence

The UK plan is modelled closely on the Australian bill. However, that ban has not so far proven effective, with teenagers widely reporting being able to work around it through attempts to circumvent age checks, use of VPNs, or ultimately choosing to migrate to more niche social media platforms currently not subject to the ban (Defend Digital Me, 2026), suggesting that implementing a similar policy will be ineffective. In the meantime, social networks continue to be allowed to host harmful content and implement harmful patterns that negatively affect all users, including teenagers.

This policy is superficially appealing, correlating teenage mental health issues with social media prevalence and serving as a political tool (Badenoch, 2026). However, evidence indicates it is unlikely to be effective. Defend Digital Me (2026) argues that the root causes of social media harm would remain unaddressed by a ban. More effective solutions would involve a legal duty of care, requiring platforms to moderate harmful content and prevent predatory behaviour - an argument that will be developed in later sections.

2. The Political Framing: “Giving Childhoods Back”

The phrasing of “Giving Childhoods Back” may sound rhetorically appealing, but it is problematic. It implies that children who can no longer access social media revert to earlier norms, such as climbing trees, riding their bikes, and unstructured play. But this imagery is clearly aimed at younger children, not the teenagers who would be primarily affected by this legislation, and seeking to apply it to teenagers is naive. Teenagers have never spent their time in this way. They seek peer connections, independence, and identity exploration, and did so long before smartphones were widespread. However, they remain barred from many adult spaces by law and, in effect, from others by a lack of disposable income. They are often unwelcome in spaces aimed at younger children, and may well have outgrown them. In recent years, many offline spaces suitable for teenagers have been hollowed out, commercialised, securitised or priced out of reach, leaving online spaces as some of the few where adolescents can gather.

As a result, teenagers often improvise their own spaces outside immediate adult control, and often semi-illicit. This has often included congregating in parks, outside shops or in other public places late in the evening. In an environment like this, they may well be exposed to illicit alcohol, tobacco and drugs, or become involved in nuisance-level crime such as petty theft or vandalism. Removing social media does not give adolescents their childhoods back in any realistic sense; it removes the infrastructure they currently use to meet the same developmental needs they have always had, forcing them to rely on improvised spaces for social interactions.

2.1 Grief, urgency, and stacked correlations

The push has also been driven in part by parents who lost children in high-profile cases where social media was potentially involved, and have subsequently campaigned for a ban. For instance, one of the high-profile advocates is Ellen Roome, the mother of Jools Sweeney, whose death she linked to a social media challenge, although the inquest into this is still ongoing at the time of writing. Another is Esther Ghey, whose daughter Brianna was murdered by a peer who had become obsessed with the idea of murder after viewing violent content online.

The voices of bereaved parents deserve compassion and attention. Their experiences demonstrate clear failures in platform governance and moderation. But grief, however understandable, is not a reason for action unless coupled with causal evidence. When Esther Ghey stated, “500 children are being referred to mental health services every day” and “97% of 12‑year‑olds own smartphones” (Mackintosh, 2026), the juxtaposition is emotionally powerful. It is not, though, conclusive: two statistics placed side by side to imply a causal relationship that has not been established. Policymakers have a responsibility to distinguish between patterns that feel compelling and patterns that are actually supported by evidence. The fact that two trends rise together does not mean one causes the other, and it is precisely in moments of public grief that evidence‑based policy becomes most important.

It is entirely understandable that parents who have suffered unimaginable loss reach for the most urgent explanations available. But urgency is not the same as causality. Mental‑health referrals have risen for many reasons — austerity, pandemic disruption, academic pressure, reduced CAMHS capacity — and smartphone ownership has risen for many reasons too. Placing these trends side by side creates the impression of a causal chain that cannot be substantiated. It is also notable that consensus is far from universal about the effectiveness of these measures among people who lost children in high-profile cases where social media was involved. Ian Russell, whose daughter Molly took her own life in 2017 as a result of depression and negative social content, has gone on record to state he believes this is the wrong approach (Rose, 2026).

3. Misdiagnosing the Problem: Harmful Content Is Not Inevitable

Campaigners argue for holding platforms accountable - a reasonable goal, since platform design choices have caused real harm. Yet the under-sixteens ban shifts responsibility onto teenagers rather than requiring platforms to address systemic issues.

If the intention is, in fact, to hold platforms to account, then they should instead be compelled to offer systems that implement effective moderation policies, that consistently flag explicit content, that do not amplify hateful content, do not promote harmful material, do not target vulnerable people, and do not expose users to unspeakable content through algorithmic incentives. These are not small start-ups; they are some of the largest companies in the world, listed on stock markets, and with compliance teams and regulatory exposure. They can be compelled to act to resolve these issues if the political will exists.

The harms seen today did not arise spontaneously, and are not inevitable. They arose due to deliberate design decisions that prioritised financial incentives over user welfare. For instance, amplification of harmful content, a lower bar to sharing misinformation, under-resourced moderation, and business models that reward outrage and vulnerability. These decisions harm not only children and teenagers, but vulnerable adults, including those with mental health conditions, cognitive difficulties and limited digital literacy.

A ban does not address these failures; it accepts harmful content as inevitable and removes pressure on platforms to change. True accountability means requiring platforms to fix the harms they create, not simply excluding one age group while leaving others unprotected.

Of particular note is this observation from the introduction to Careless People by Sarah Wynn-Williams, made by Naomi Alderman:

Or one might reflect on the fact that, as (Nick) Clegg pointed out, for some “vulnerable teens”, using Instagram causes mental health problems. Meta’s own internal slides, as reported in The Washington Post, say “We make body issues worse for one in three teen girls.” Reading stark words like that, a person could sit and wonder why a company that discovered that kind of effect would not want to act swiftly and decisively to bring that damage to a halt. What kind of ethical husk would a person have to be to put profit above the mental health of children? What on earth would a person like that think the purpose of their one human life is? Do they think it’s not to attempt to create something of lasting value, to use their money and power and influence to help the world in some meaningful way?

Banning under-sixteens would only resolve this issue in part. It would create a cliff-edge where sixteen-year-olds would suddenly be allowed in an environment that would be no less toxic than it is now. It is clearly naive to think these body issues do not affect girls of sixteen and over.

3.1 The ban assumes teenagers are uniquely vulnerable

It has often been claimed that teenagers are uniquely vulnerable. However, there is no evidence to support this claim.

Policymakers often justify the ban by appealing to adolescent brain immaturity, particularly the development of the frontal lobe. However, contemporary developmental neuroscience does not support the idea of a single, biologically meaningful age threshold. Adolescent brain development is asynchronous and non‑linear, with different systems maturing at different rates throughout the teenage years and early adulthood. As a recent policy brief notes, “developmental sensitivity fluctuates throughout the entirety of adolescence, suggesting no biologically optimal age at which unrestricted social media exposure becomes appropriate” (Frontiers in Digital Health, 2026). This directly contradicts the notion that 16 represents a meaningful neurological boundary. Other research similarly emphasises that regulatory systems continue developing well into the mid-twenties (Policy Insights from the Behavioral and Brain Sciences, 2026). If policymakers genuinely believed frontal‑lobe immaturity justified exclusion, the logical threshold would be twenty‑five, not sixteen.

It is also often claimed that teenagers are not capable of recognising potential harms on the Internet. However, in practice teenagers are often significantly more digitally literate than their parents. More of their interactions with peers take place on social media platforms, and they have grown up with them present in a way older people have not unless their background has exposed them to the Web more deeply. Many may also be studying subjects such as Media Studies, which explicitly teach how to treat media sources sceptically. As such, adolescents are actually very able to develop the sort of “street smarts” that keep them safe, and often share details among themselves of interactions to watch out for. It is also common for them to support older family members in these matters, suggesting this claim is not supported by evidence.

What is clear, however, is that adolescence is a uniquely social time of life, when peer networks become much more significant factors in a young person’s life. Young people are immersed in social media because it is where those peer networks are present. Before the advent of social media, adolescents were still among the heaviest phone users, but used SMS to maintain complex peer networks. Before then, the only solution, barring going round someone’s house and running the risk they might be out, was to phone their landline and potentially deal with someone else picking up the phone. It is therefore unsurprising that adolescents are heavy users of social media, because it provides an always-on, asynchronous channel to communicate with friends, make new friends outside of their established circle, and participate in dedicated communication channels for hobbies and interests. Adolescents are excluded from many adult channels for socialisation, such as pubs and nightclubs (due to the presence of alcohol), or cafes (due to lack of funds), and third spaces such as youth clubs are potentially vulnerable to withdrawal of funding. Social media provides a free alternative for communicating with friends.

3.2 The evidence for adult harm is at least as strong

The parents of people who are currently teenagers occupy a wide age range, but few are young enough to have grown up immersed in social media to the extent that modern teenagers are. Therefore, unless they have had that sort of deep exposure through other means such as working in a related field, media literacy training or being an early adopter of technology who was immersed in it for leisure, they very often have a worse understanding of the potential dangers than their children do. When social media first became prominent, adults often became fixated on headline fears like grooming, which were actually fairly well understood risks that had solid controls in place already and were already covered by digital safety lessons in schools and other communications around “stranger danger” that have existed for decades. However, they missed the everyday, high-priority risks adolescents faced, such as inadvertent oversharing of information, boundary-crossing and misread signals. Adults who only use the web transactionally (such as Facebook Marketplace, Mumsnet and online shopping) do not see and appreciate the subtler, everyday risks that emerge from dense social use and about which adolescents are often far better informed.

Adults themselves are also at risk from potentially nefarious uses of social media that adolescents generally are able to identify. Their lack of immersion in digital environments and lack of digital literacy training often means they are less likely to be able to identify misinformation, dog whistles, behaviour indicative of bots and algorithmic amplification and other potentially dangerous patterns. For those with dementia, cognitive decline or learning difficulties, these risks are amplified further. These can leave adults at risk from scams, political radicalisation and manipulation that most adolescents have learned to recognise.

Restricting only under‑sixteens clearly excludes the group with the highest digital literacy while leaving the most vulnerable adults exposed, strongly supporting the idea that it is the wrong approach.

3.3 Young people do not believe it will be effective and do not want it

It is also notable that when young people are asked what they believe would make them safer online, they do not call for bans or age cut-offs. Multiple consultations with young people show that young people want better moderation, stronger reporting tools, and platforms that take responsibility for harmful content (Defend Digital Me, 2026). They see bans as ineffective, easily circumvented, and likely to push harm into less regulated spaces. In other words, the measures young people themselves identify as effective are precisely those recommended by domain experts and which a ban on under-sixteens avoids.

Some campaigners have sought to portray children opposed to a ban as being “addicted” and displaying a knee-jerk response (Lewis, 2026); however, this assertion is not supported by the evidence. Others have sought to portray domain experts opposed to a ban as having a vested interest. Given that those domain experts include campaigners who have consistently opposed social networks on privacy and behavioural incentives, this seems absurd. These failures of diagnosis lead directly to failures of implementation, most notably the assumption that age verification can reliably enforce a ban.

4. Practical Failure: Age Verification Does Not Work

Age verification online is often treated as a solved problem, yet the Australian ban has already demonstrated that this is not the case. Current methods are either easy to bypass, dangerously invasive, or impossible to deploy at national scale. The UK lacks any digital identity system capable of safe, effective age checks. Enforcing such verification would introduce new risks—from identity theft to mass data collection on children—without reliably restricting access. In reality, any policy that relies on perfect age verification is built on wishful thinking. Unless policymakers commit to a massive infrastructure project—which would both delay and complicate implementation—such a ban cannot succeed.

The proposed under‑sixteens ban depends on effective age verification, yet age verification is the part of the policy least grounded in reality. It is also the part most consistently flagged by privacy advocates, digital‑rights groups, and domain experts as unworkable, intrusive, and unsafe — yet most often hand‑waved away by campaigners and policymakers. Every available method is either trivially circumvented, dangerously privacy‑invasive, or technically infeasible at scale. The UK has no national digital identity infrastructure, no safe attribute‑based verification system, and no way to prevent determined teenagers from bypassing checks using shared devices, borrowed accounts, or simple workarounds. In practice, the ban relies on a mechanism that cannot deliver what the policy requires. Age verification is not merely imperfect; it is structurally incapable of achieving the level of certainty the ban assumes.

4.1 Age verification is unreliable

It is naive to assume people will not ignore age limits when it suits them. Parents have long had an inconsistent relationship with age restrictions on other products such as films and video games, with many turning a blind eye to their children bypassing age checks, or even being actively involved in doing so.

Age verification is also far from foolproof. There are many anecdotes from the Australian ban on children borrowing credit cards, using video footage of adult family members, or drawing a moustache on their faces to fool automated systems - anyone with a passing knowledge of computer hacking knows that time-rich and cash-poor teenagers are often extremely creative in terms of bypassing systems adults think of as secure.

In March 2026, 438 security and privacy researchers signed a joint statement calling for a moratorium on age-assurance deployment, warning that the checks are ineffective, invade users’ privacy, and build infrastructure that could be repurposed for surveillance.

In August 2026, a report from Australia’s eSafety Commissioner (Given, 2026) found that more than 80% of children are still using social media three months after the under‑16 ban took effect, with daily usage and account ownership barely changed and platforms failing to implement effective age‑assurance measures. Parents became less aware of their children’s social media use, children migrated to other platforms, such as Reddit and Pinterest, and little or no change was reported in offline activities. This early evidence strongly reinforces the argument that bans are ineffective, easily circumvented, and distract from the need for platform‑level regulation. It also strongly suggests that the similar UK plan will also be ineffective.

4.2 Age verification creates new risks: identity theft and data exposure

Age verification as implemented in the UK scenario is also potentially problematic. Digital verification of age is fundamentally not like the situation where alcohol is purchased, where the shop staff ask to see ID, a card is handed over, reviewed, and then returned to the owner. It involves making an electronic copy of a document, or the user’s face, which is then stored by the third party responsible for carrying out the age verification for an indefinite period of time. This copy must be retained securely and then disposed of correctly once no longer required.

Any third party that stores this kind of data becomes a potential high-value target for attackers. If, for instance, scans of documents were held on an improperly secured Amazon S3 bucket, the personal documentation of tens of thousands of people could easily be discovered to have been taken by attackers and used for identity theft purposes, possibly months or even years after the fact. This scenario is one of the most common forms of data breach in the real world.

Privacy advocates have also questioned the wisdom of encouraging children to normalise the idea of posting video footage of themselves on request. Schools are actively being warned about taking photos of pupils offline because of AI-generated abuse imagery, but simultaneously children are asked to hand over a high-resolution image of their face for access to social media.

This could be achieved more effectively with sufficient political will. Estonia has a digital identity infrastructure which allows zero-knowledge identity verification and yes/no verification of a single attribute. In other words, a service can ask “Is this person over eighteen?” and the system will disclose “Yes” without confirming any non-essential details. However, the UK does not have that sort of infrastructure in place and has not demonstrated the will to implement it.

In short, any ban which necessitates unsafe means of age verification is not a realistic safety measure. It is of questionable utility in itself, and exposes a very serious attack surface for users’ personal data.

5. New Harms Created by the Ban

These measures would cause other harms as well. Adolescents might continue to use these platforms via VPNs, preserving the harm while bypassing national jurisdiction. The legislation only covers the most high-profile platforms, encouraging adolescents to migrate to more fringe platforms, which may be an unknown quantity and could be more dangerous, with far more limited moderation policies.

Such a ban would also harm intergenerational digital literacy, as social media is now a fixture of adult life. Restricting adolescent access reduces their ability to develop essential digital skills, an argument already made in earlier sections.

5.1 Isolation of vulnerable young people

Removal of access to social media would cause other social harms. Teenagers who are neurodivergent, LGBTQ, have a disability, are from ethnic minorities, isolated in rural areas, bullied, or have niche interests often rely on social media to talk to people like them and form communities. Teenagers who are forced to relocate or isolated by family circumstances or caring responsibilities often rely on social media to remain in touch with friends. Without social media, it is unlikely those bonds could be maintained, leaving many adolescents who are otherwise marginalised at serious potential risk of social isolation.

Severe social isolation is strongly associated with depression, anxiety, self-harm, suicidal ideation, reduced academic engagement, impaired identity formation, and long-term mental health consequences. It leaves young people with unmet social needs that can be exploited through predatory behaviour such as grooming or radicalisation. It is far more immediately harmful than the typical online risks adults fixate on.

5.2 Misleading product‑safety analogies (“take it off the road”)

Ellen Roome has argued that “If this were a product, it would be taken off the road” (Mackintosh, 2026). However, this analogy is flawed.

A social network is not a discrete consumer product. It is an ecosystem, made up of other users, relationships, and infrastructure. It is not a product you purchase, but an environment you choose to inhabit. The hazards arising from it do not derive directly from its technical implementation, but from how the users are incentivised to use it by aspects of its design, and the presence or absence of young people will not change that. If a social network is implemented responsibly, in a way that encourages healthy usage and does not encourage unsafe behaviours, it is no more unsafe in principle than any other channel for socialisation, such as text messages or telephone calls.

Early on, Facebook displayed a relatively straightforward list of things that had been posted by friends and pages that the user followed. While not without problems, this model was relatively benign compared to its modern implementation. The problems people were mostly concerned about at the time in relation to children’s access to social media were around the potential risk of grooming by sex offenders. They were generally handled the right way, through regulation, with CEOP, mandatory reporting tools, online resources, moderation requirements, police liaison units and other appropriate measures. A slightly less obvious, but probably more common, potential problem was that it provided a channel for bullying that is more widely accessible, so that children might be bullied by peers outside of school hours. However, social media merely facilitates bullying rather than creating it, and arguably provides more means for preventing bullying compared to text messages, phone calls, written messages and verbal communication. Direct message systems provide a full audit trail of a conversation, with a record of all communications sent, as well as tools to block and report abusive communications.

However, since then Facebook has noticeably degraded the user experience in ways that encourage unsafe behaviours, as noted in Enshittification (Doctorow, 2025). Advertisers were sold unprecedentedly accurate profiles of users, allowing them to target them for products and services with a far greater degree of confidence in a sale than ever before. In short, it did not become more harmful because teenagers changed, but because its business model changed and the implementation changed to reflect that. It therefore follows that if the will exists, these changes could in principle be reversed, and a healthier social media ecosystem could be created.

A faster-paced and more pronounced example of this same phenomenon took place when Elon Musk purchased Twitter and rebranded it as X. Soon after the purchase, Musk gutted moderation teams, weakened moderation policies, reinstated many abusive accounts which had been banned, dismantled the prior identity verification system in favour of a paid scheme anyone could join, amended algorithmic amplification to favour right wing views, and shifted incentives towards engagement at any cost. The platform quickly became much more unpleasant, leading to a feedback loop pushing the general political views on it to the right as those who were previously on platforms like Parler migrated to the platform, and those on the left who were seeing more combative interactions and less meaningful engagement decided to look elsewhere. Twitter did not become unsafe because of who used it, but due to the dismantling of the safety measures by its custodians.

In the early days of the Web, it was quite common to make an analogy with the Wild West in terms of legal protections. While there are undoubtedly parts of the Web which remain like that, social networks are fundamentally not one of them. They are owned and operated by huge companies, often listed on various stock markets, and with permanent staff and legal offices, and it is entirely possible to hold them to account through legal and regulatory means such as the European Union’s Digital Services Act. Failure to do so may be more likely to suggest a lack of understanding by elements within government, a lack of political will to resolve these issues correctly, or the result of active lobbying by the services in question.

The analogy with taking a car off the road also suggests a potential comparison with Unsafe at Any Speed (Nader, 1965), which held the automotive industry in the US accountable for systemic failures in safety, such as disregarding safety concerns, and lobbying against mandatory implementation of safety features. After this, the United States government established various departments with briefs that related to vehicular safety. It is also notable that Facebook were also criticised in Careless People (Wynn-Williams, 2025), and responded with legal threats that bordered on the absurd - a comparison with the attempts at character assassination of Nader by GM is not wholly unwarranted. Nader did not argue that teenagers should be banned from driving, but that manufacturers should be obliged to make cars safer. The same logic should be applied to social media. Together, Careless People and Enshittification may in time play a similar role in respect of the harms present on social media comparable to that begun by Nader.

In summary, the analogy of a social network with a physical product supports the idea of implementing better regulation of that network more clearly than banning under-sixteens. Every example- from early to modern Facebook, Twitter to X, and the automotive industry - demonstrates the same pattern. These harms arise from governance failures, and these should be resolved through regulation, not prohibition. This holds those responsible for poor design decisions to account, and also protects all users, not a subset of them. Social media is unsafe for teenagers not because they are allowed to use it, but because platforms have been allowed to degrade their own safety architecture. Social media is not intrinsically harmful, but particular implementations can be.

5.3 Social media is not a separate sphere, but a discrete channel for existing relationships

It is also important to recognise that social media is not regarded by adolescents as a separate sphere. The overwhelming majority of their online interactions are with people they know in real life - classmates, friends from current or previous schools, or peers from clubs or groups. Social media is simply a channel for maintaining the same relationships that exist offline. Removing access therefore does not remove a harmful environment; it removes part of the infrastructure through which teenagers sustain their existing friendships. For adolescents who are already isolated, such as those who have been out of school for an extended period, this loss would be not merely inconvenient but potentially devastating.

6. A Better Alternative: Literacy, Regulation, and Shared Responsibility

Given that an under-sixteen ban appears to be problematic, how else can we resolve these issues? The following measures would appear to be supported by the available evidence as practical, achievable and effective.

6.1 Platform regulation and accountability.

Given that we have argued that the harms of social media are far more widespread than under-sixteens, and arise from deliberate decisions by the platforms, the most important factor is clearly to prevent those harms from being permitted in the first place. To that end, I would suggest instituting a mandatory code of conduct for social media sites that wish to operate in the UK, backed by regulatory powers. The exact shape of this code of conduct would be a matter for the implementers to decide. However, as a starting point, I might suggest:

• Zero tolerance for certain hateful behaviours and harassment of minors, perhaps - immediate ban for those users and measures to prevent them from immediately creating a new account
• A minimum standard for blocking and reporting - perhaps specify what a block must do to ensure it is effective, as well as a statutory minimum for what the reporting procedure must do and how it must be evaluated. Bluesky implements a highly effective block system that could be a model for this.
• Considering the institution of an independent body to whom reports can be escalated if the response is unsatisfactory
• Banning certain addictive patterns

This would allow the removal of harmful aspects of social networks, making them a more appropriate place for young people to be present. Failure to abide by this standard would result in disciplinary action, up to and including blocking sites at the national level.

The 2026 California court case was built around the core allegation that Meta’s platforms are structurally addictive — not accidentally, but by design. Infinite scroll, algorithmic amplification, engagement‑driven ranking, and behavioural nudges were all central to the complaint (Griffin, 2026). These are the same mechanisms identified earlier as the true drivers of harm. The case therefore demonstrates that the fundamental problem is platform design, not teenage access. A ban cannot reach these systems; regulation can.

The measures Meta agreed to — time limits, notification blocks, hidden likes, optional non‑algorithmic feeds — are behavioural nudges, not structural reforms. They resemble the “timeouts” used in the online gambling industry: highly visible, easy to advertise, and largely symbolic. They do not alter the underlying systems that generate harm. Meta has not been required to change its algorithms, remove infinite scroll, reduce engagement optimisation, or redesign its content‑discovery pathways. These are surface‑level mitigations, not the deep reforms experts have long argued for.

The financial penalty — up to US$18 billion — is unprecedented (Hays and Rahman-Jones, 2026). It signals that regulators are finally willing to impose material consequences for design‑driven harm. But fines alone do not change platform architecture. They create pressure, not safety. Without structural obligations, the underlying systems remain intact. In addition, the fines were modest compared to those that could have been imposed (Stokel-Walker, 2026).

The lawsuit went after the actual mechanisms of harm: addictive design, algorithmic amplification, engagement‑driven ranking, and internal decision‑making that prioritised profit over safety. A ban, by contrast, targets only user behaviour — specifically, teenage access. It does nothing to change the systems that harm adults, vulnerable users, or teenagers once they turn sixteen. The Meta case therefore demonstrates that platform‑level accountability reaches the root cause, whereas age‑based bans do not.

6.2 Digital literacy training measures for teenagers and adults.

Parents often came to social media later in life than their children did, and many of the more harmful aspects of them have crept up on people. It is therefore worth considering whether both parents and children would benefit from digital literacy measures.

In 2009, Channel 4 commissioned an interactive game called Smokescreen to teach 14–16‑year‑olds about online privacy, stalking, phishing, and the risks of oversharing. As the designers put it, “Smokescreen aims to make young people think about solutions to online privacy and security as the risks and threats could actually be real.” It simulated the internet through storylines that showed how ordinary people can misread online signals or cross boundaries without understanding the consequences. One plot followed a boy with amnesia trying to understand why a girl was running a podcast about him — only to discover she did not know him at all and was simply obsessed with him. Another showed a young man who had never met a girl he’d friended on “Fakebook”, but used her posts to track her movements and pursue her on a night out. She was horrified when she realised he was not just a random man in the club. These were not morality tales about “reckless teens”; they were interactive demonstrations of how anyone can get into trouble online without realising it (Woods, D. and Scott, E.S, 2012).

At the time, rhetoric around potential harms from social media tended to focus on grooming. Smokescreen demonstrated that there were many other potential dangers, most of which were not innate to any particular age group. These issues were just as much a threat to adults, and the characters portrayed were primarily aged 16-18, so would not be covered by the proposed social media ban.

An updated workshop along similar lines, run by schools, could be effective, with parents and children working together to navigate a mocked-up social media network, identify potentially risky situations, find fake content, and see the consequences of ill-advised behaviour play out in a safe environment.

7. Conclusion: Symbolic Politics Will Not Protect Children

A ban on under-sixteens is largely symbolic, with little expected impact on actual harms. As argued throughout, it risks depriving digitally literate adolescents of vital social channels while failing to address the core governance failures that create those harms. Real safety hinges on holding platforms accountable and prioritising systemic solutions rather than symbolic politics. Policymakers should resist symbolic measures and instead pursue evidence-based regulation that addresses the structural causes of harm.

Similarly, the US$18bn settlement and the measures agreed in response to the California court case, while more effectively targeted at the root causes of the problem, have not resulted in fundamental changes in the addictive nature of the platform, an issue which affects adults no less than children. A useful comparison could be drawn with the online gambling industry, which has embraced measures such as timeouts to the point that it is mentioned prominently in advertising, but has not fundamentally changed the addictive nature of its business model.