Ofcom’s new report asks whether online safety is reaching children
Ofcom’s first statutory report on content harmful to children finds no statistically significant overall change in children’s reported exposure during the first nine months after the UK’s child safety duties took effect. It also recommends stronger statutory treatment of body stigma and depression content. The report raises an outcome question, but it does not establish that the duties have failed.
What Ofcom found
Published on 22 October under section 63 of the Online Safety Act, the report reviews the incidence and severity of content harmful to children. Most children surveyed said they had encountered at least one harmful item in a four week period. In the March to April 2026 survey wave, 42% of 11 to 17 year olds reported encountering body stigma content and 23% reported encountering depression content in the preceding four weeks. Ofcom recommends moving these two kinds of content into the Act’s priority harmful content category, so that covered services would face clearer duties concerning them if the government makes that change. The recommendation is not itself a change in law.
What remains unresolved
The headline finding needs careful interpretation. Ofcom compared children’s reported experiences before and after the duties began, but this report does not evaluate whether individual services complied with their duties or whether particular safety measures caused the observed trend. No statistically significant overall change is not proof that every measure failed, nor does it mean no child benefited.
- The exposure findings draw on children’s recalled experiences. Further evidence should test what children actually encounter, how often, and through which product features.
- Overall exposure can obscure differences by age, vulnerability, type of content and service. A child may encounter repeated harm even when a platform reports that its controls are in place.
- Ofcom’s proposed priority classification for body stigma and depression content needs government action before it becomes a statutory change.
- The report covers specified harmful content; it is not an evaluation of harmful contact, illegal content, AI functionality, or the effectiveness of the children’s codes of practice.
- Policy evaluation should examine what happens as children move among services, including recommendations, search, messaging and signed out experiences.
The Child-Friendly Internet perspective
For the Child-Friendly Internet Initiative, the test is the child’s experience, not only a service’s stated compliance. Participating services and networks could use a trustworthy, purpose limited indication that a child appropriate response is needed, without each service collecting the child’s name or identity document. Each provider would still have to design, moderate and test its own service, and regulators would need independent measures of actual exposure and wellbeing. Interoperable assurance is a concept to evaluate, not an outcome established by this Ofcom report or a substitute for enforceable service duties.
- Measure changes in harmful exposure and wellbeing across the services children actually use.
- Test whether protective defaults work for inaccurate, borrowed and shared accounts as well as correctly classified accounts.
- Minimise identity disclosure and avoid a persistent identifier that could track a child across services.
- Preserve beneficial access, participation and increasing autonomy while reducing harmful exposure.
The central question
Are children experiencing a safer Internet, or are we mainly counting the safeguards that individual services say they have introduced?