A new study from the Youth Endowment Fund (YEF) says a third of teenagers surveyed in England and Wales have found a way around online age checks, and nearly half say they would find ways to keep using social media if laws restricting them came into effect.
The UK's Online Safety Act (OSA) has required highly effective age assurance for pornography and certain other content harmful to children since July 2025, including on social media platforms. The government also plans to introduce restrictions preventing under-16s from accessing certain social media services in spring 2027.
According to the YEF survey, in total, 32 percent of teens surveyed have bypassed age checks. Among those who had bypassed age checks, 25 percent accessed sites showing nudity or sexually explicit content, 18 percent accessed fictionalized violence, and 15 percent visited sites promoting gambling or betting.
The most common methods for bypass included virtual private networks (VPNs, 12 percent) or using websites without age checks (nine percent).
In its release, the Youth Endowment Fund says the findings “raise fresh questions about how effectively online protections can keep children safe from harmful and violent content.”
However, they can also be interpreted through the lens of several basic truths.
Kids will find sex, but social media is something else
The first is that, in any world, under almost any circumstances, some part of the young population will find ways to satisfy their curiosity about sex. A quarter of kids skipping age checks to look at porn sounds about right, and is no reason to believe regulations aren’t working. Because what we don’t get in the YEF’s data is the number of teens not looking at porn because of age checks – arguably a much more significant (if tricky to nail down) number.
The second is that responsibility for protecting children cannot rest on age checks alone. Platform design and content moderation also matter. The headline finding on online violence is that “six in ten teenagers (61 percent) said they had seen real-world violence on social media in the past year.”
The number is down from 70 percent in each of the previous two years. The decline coincides with the OSA's child safety duties, although YEF says its research cannot establish whether the legislation is responsible. But violent content is still rife, including “weapon-related content.”
Platforms’ model incentivizes sensational content
The findings point to structural and philosophical issues that go far beyond how effectively tools are blocking kids from seeing online porn. In terms of violence, on the one hand, the survey included in this category “content such as videos of fights and threats to beat up another child.”
Sadly, these are things many kids encounter in physical space, as well.
But measure this against the finding that teens say seeing violence online changes how they behave. For 22 percent, it instills fear and causes them to leave home less often. Eight percent feel more aggressive.
Or, measure it against the emerging correlation between
AI and online violence and abuse. The survey says teenagers who had perpetrated violence were more likely to have created or shared fictional violent content that can be used to intimidate, humiliate or provoke others.
Or, again, against the finding that almost half of teenagers believe spending less time on social media would improve their family relationships (49 percent) and mental health (45 percent).
What emerges as the online safety era continues to unfold is a problem that is less about policing access than it is about corporate responsibility, regulation and technology. The content on social media is a symptom of the root problem – which is the design, structure, business model and operational ethics of social media platforms. Views, clicks, attention and data are still those platforms’ most valuable currency, and as long as violence is viral, they will be incentivized to showcase it.
Jeremy Miles, CEO of the Youth Endowment Fund, understands this. “Tech companies must do more to protect children from harmful and illegal content,” he says. “We have laws in place to hold them accountable, and those laws must make a difference in what children see online.”
“We need stronger protections from platforms, alongside education to help children understand the risks of what they see and share, and the consequences of using social media to harm others.”
Age checks exist; now law must hold platforms accountable
On an executional level, physical world analogues still help to put the numbers in context. Andy Lulham, COO of UK age assurance provider
VerifyMy, says “the findings show that weak age checks without robust anti-spoofing measures are no match for resourceful teenagers,” and that “poorly implemented age gates, such as facial age estimation with no
liveness checks, are easy to get past.”
But he notes that “many of the workarounds young people describe involve an adult handing over their phone or helping them through a facial age check. This is not dissimilar to buying alcohol for someone underage.” Seen this way, the issue is not that the law is not appropriate, but that some people will not follow it. The answer here is a coherent and consequential
enforcement regime, and an approach that loops in devices, as well. Biometrics are a potential solution.
Lulham notes that “the
technology to protect young people online already exists, and it works. Robust methods such as ID checks with face matching or
email-based age checks are highly effective and have proven much harder to beat. But technology and regulation can only go so far.”
“Young people deserve the same shared responsibility online that we expect offline, and that starts with better education for adults about the harms children could face otherwise.” Which is to say, people should know that buying booze for kids is, generally, a bad idea that can get you in trouble.
At least part of this is acknowledging that, on a high level, letting social media platforms become so ubiquitous and unfettered by basic social obligations or regulations has been a grand societal mistake. Like booze, social media can be a great numbing escape; but it’s addictive, unhealthful and not for kids. Regulators can’t fix online violence and sex, but they can fix the engines and business models that make those things into assets. And, as discussions about a
duty of care model develop, they appear to be trying.