UK Online Safety Rules Fail to Cut Harm for Children

Britain's landmark Online Safety Rules have not made the internet measurably safer for children, according to the regulator charged with enforcing them. In its first statutory review of content harmful to children, Ofcom found no significant overall change in how often young people encounter harmful material since platforms' child safety duties began in July 2025, a result the regulator itself called concerning.

Most children in the United Kingdom told researchers they had seen at least one kind of harmful content within a four-week window, according to the review, which drew on surveys of children aged eight to seventeen. The findings land more than a year after the duties took effect, and they raise an awkward question for a law sold as a turning point in child protection: if the rules are working, why does the feed look the same?

What the review actually found

The headline finding is blunt. At a total level, there was no significant change in the share of children encountering harmful content over the first nine months after the duties came into force, according to Ofcom. The regulator also said it found no evidence that the severity of harm had changed since the flagship requirements began, with unsafe material often leaving children frightened, confused and upset.

Some categories of harmful material remain strikingly common. In surveys from March and April 2026, almost half of children aged eleven to seventeen said they had seen hate content in the previous four weeks, just over four in ten reported posts shaming body image, and nearly a quarter reported low mood posts, according to the review. Across several categories, children most often said they came across harmful posts through personalised feeds, though comments, direct messages, group chats and online games also played a part.

Ofcom pointed repeatedly to recommendation systems as the engine of the problem. Personalised feeds powered by for-you algorithms were repeatedly found at the root of harmful material still reaching children, according to the review, because the systems select content based on what a young user has previously viewed or engaged with. That matters for anyone who has watched a feed drift from a hobby video into something darker within a few swipes: the feed is not neutral, and the regulator now says so in its own evidence.

What Ofcom wants changed next

Rather than declaring the Online Safety Rules a failure, the regulator recommended widening them. Ofcom advised that body stigma content and depression content should be added to the category of priority content that is harmful to children, which would require services to protect children in age groups judged at risk from encountering it. The advice matters because repeated and cumulative exposure to such material can create a material risk of significant harm, including poorer mental health, body dissatisfaction and barriers to seeking help, according to the review.

The recommendation has already split opinion. Campaigners called the lack of progress staggering, according to press coverage of the review, arguing that children were promised protection and are still waiting for it. Critics of the law go further, saying the regime has age-gated a wide range of material for adults, from war news to art images, while failing to move the numbers for children. Supporters of tougher enforcement counter that the duties are still young, that the review does not judge whether individual platforms are complying, and that the next step should be stronger action on feeds rather than retreat.

What happens next sits with ministers. If the Secretary of State takes the recommendation forward, protections against body stigma and depression content would be written into the law itself. Ofcom also said it will soon publish an assessment of what highly effective age checks look like, as the government weighs a possible ban on social media for children under sixteen. The experience of Australia, where young users have found ways around a similar ban, is already being cited by sceptics as a warning about what enforcement alone can achieve.

For young people, the practical takeaway is less about legislation and more about the feed itself. The review shows that harmful material is not mainly arriving through searches for it; it is being served up by recommendation systems, often in high volumes. Until those systems change, the safest habit remains an old-fashioned one: curate follows deliberately, use mute and block tools without hesitation, and treat a feed that keeps serving distressing content as a signal to step away rather than scroll on. The regulator has now admitted the numbers have not moved. The test of the next nine months is whether the feeds do.

For more on how platforms shape what young people see, read our coverage of Canada's push for safer social media and UNICEF's findings on online child abuse, and see how Britain is pairing tech rules with a major new AI safety investment. The full review is published by Ofcom, with analysis from Reason and Digital Watch.