Proposals to restrict or ban young people from social media and algorithmically driven media platforms are becoming increasingly common around the world. The most far-reaching example is Australia, which has legislated to prevent under-16s from holding accounts on many social-media platforms.
Elsewhere, governments are moving in a similar direction. France and Denmark are considering under-15 restrictions. Greece has announced a ban for under-15s from 2027. Indonesia has imposed under-16 restrictions on designated high-risk platforms. Malaysia has signalled comparable plans.
The UK government is consulting on minimum age limits and restrictions on “addictive” design features such as infinite scroll and autoplay. The European Union is developing a common age-verification system.
This amounts to a patchwork of different approaches, but they are united by a common concern: that the current social media model may be damaging to young people.
Concerns
Much of the left has largely rejected such moves. Almost all left commentators emphasise the social context shaping young people’s lives: economic precarity, educational pressure, anxiety about climate change, and the long-term effects of austerity on youth services and mental-health provision. Some argue that the focus on social media exaggerates its role, masking deeper structural causes of distress. Others go further, suggesting that moral panic about social media risks becoming a vehicle for an authoritarian suppression of young people’s access to information and channels of political communication.
These are valid concerns, but in my view they tend to be a substitute for asking further questions. First, is the current ubiquity of largely unregulated social media contributing to the deterioration of young people’s mental health? Second, if so, can state intervention reduce those harms without imposing disproportionate costs on young people’s freedom and agency?
There is now substantial evidence of a problem. The expansion of smartphones and social media from the early 2010s coincides with a marked rise in reported mental-health difficulties among young people. In England, the proportion of 17–19-year-olds with a probable mental disorder rose from 10.1% in 2017 to 17.7% in 2020, and higher still by 2022. The deterioration accelerated through the Covid-19 pandemic, but only by intensifying existing trends. Sweden, where there was no Covid lockdown and where schooling continued, had less of a Covid spike in youth mental ill-health, but there too young people’s mental health has continued to decline longer-term.
Social Environment
Of course the broader social environment matters. In England, local authority spending on youth services has fallen dramatically since 2010, by around 70% in real terms, with the closure of more than 1,000 youth centres. These cuts have reduced access to in-person spaces where young people can meet, organise and find support. However, similar declines in adolescent mental health are seen in European states where youth services have not been so drastically cut, suggesting that austerity alone is unlikely to be the complete explanation. Economic and social pressures are relevant, but cannot carry the full explanatory weight.
What has changed most clearly is the structure of young people’s social lives. Platforms such as Instagram and TikTok do not simply host communication; they actively organise it. Social interaction becomes continuous, public and quantified. Status is measured in likes, shares and followers. Content is filtered not chronologically but by engagement. It is plausible that the effects of the pandemic were amplified by this shift: in-person interaction was replaced by online interaction, and those patterns have persisted.
The potential consequences are well documented. For many girls, the tendency is for social media can intensify comparison and body-image anxiety. For boys, it leans toward creating pathways toward misogynistic or extreme content through recommendation systems. More broadly, high levels of use are associated with poor sleep, anxiety and depression, and may displace other activities linked to wellbeing.
These effects are not accidental. The dominant social-media model is built around maximising engagement. Features such as infinite scroll, autoplay and persistent notifications are designed to retain attention. This does not mean that social media is the sole cause of the current mental-health crisis. But it is strong evidence that it forms part of the problem. And not a trivial part.
If that is the case, then the question becomes one of response. Market self-regulation has so far proved limited. Parental control is difficult where social media are embedded in peer culture and technology is so widely available. Individual self-discipline is constrained in environments explicitly designed to capture attention. In practice, the only actor capable of imposing system-wide constraints on large platforms is the state.
That does not mean that all forms of regulation are equally desirable. An outright ban on younger teenagers using social media is a blunt instrument. Critics argue that such bans will be widely circumvented, and that they risk creating a false sense of security. The Molly Rose Foundation have warned that parents may assume children are safe simply because formal restrictions exist, while harmful content remains accessible through other channels.
However, the fact that regulation is imperfect does not mean it is ineffective. Age limits on alcohol and tobacco are routinely bypassed, yet they still shape behaviour and establish social norms. The same may be true of digital regulation. Even partial enforcement can alter expectations and reduce exposure at scale.
Platforms
At the same time, regulation needs to address the behaviour of platforms themselves. Age restrictions are of limited value if companies do not take meaningful steps to enforce them. There is also growing concern about the ways in which recommendation systems can direct users, including minors, toward increasingly harmful content once initial engagement is established.
Elon Musk’s X is the worse offender here. A recent study has shown that a simulated 13 year user making a few curious sex-related searches is algorithmically fed prompts to lead them to hard-core, bestial, and under-age pornography. But other platforms are little better. Data Protection regulation has created a de-facto “digital age of consent” of 13 years old, but the majority of young people have social media accounts before that age. The platforms do not age-verify. They have the ability to spot age-related use and block it. They don’t.
The UK’s Online Safety Act 2023 was intended to address some of these issues by imposing duties of care on platforms. However, its impact so far has been limited and public willingness to challenge big tech companies weak.
Pressure for change may also come through the courts. In 2026, a California jury found Meta Platforms and Google liable for harm, awarding $6 million in damages after finding that their conduct contributed to a young woman’s mental-health deterioration. There are thousands more cases in the pipeline and this may push platforms toward reform. But state action would be preferable to the big tech companies balancing their profits against legal claims of harm.
It would be wrong, however, to treat social media as purely harmful. For many isolated or marginalised young people, particularly as those questioning their sexuality and gender identity, it provides meaningful forms of connection and support. Any regulatory approach must take these benefits into account. Of course, we should campaign for liberatory spaces in schools and local neighbourhoods, but we should not casually allow existing support mechanisms to be disrupted.
A socialist agenda for change most include serious response to the issue must include rebuilding youth services, expanding public space and creating environments in which young people can meet and organise offline. It should also involve engaging young people directly in the development of policy, rather than imposing top-down solutions that risk being experienced as, at best, paternalistic.
The challenge, then, is not whether to regulate social media, but how to do so in a way that addresses genuine harms while preserving the positive aspects of digital life. That requires moving beyond both uncritical acceptance of the status quo and reflex opposition to state intervention. The scale of the issue demands a serious and balanced response.
Comments?We welcome replies, corrections and debate pieces. Contact us