UK regulator sets user bans over child sexual abuse material


Ofcom’s package also expands detection of known abuse material and requires some services to find previously unknown content; the regulator expects the duties to take effect in early 2027 if Parliament approves them.
UK media regulator Ofcom announced on 16 November 2026 that online services would have to ban users who generate, upload or share child sexual abuse material. The duties would apply to social media platforms, file-sharing services, pornography sites and other services.
The package is part of the rollout of Britain’s Online Safety Act, under which Ofcom can fine companies up to 10 percent of global revenue. Ofcom expects the new duties to take effect in early 2027, subject to parliamentary approval.
A wide range of services, including pornography sites, would have to use hash-matching to find and remove known child sexual abuse material, Ofcom said. The technology compares uploaded images and videos with databases of known illegal content so platforms can identify and remove matches automatically.
Some services would also have to use automated tools to identify previously unknown material, CNA said. Unlike hash-matching, that requirement concerns content that is not already in a database of known material.
Platforms with duties to prevent grooming would have to protect children’s accounts by default. Certain platforms would also have to take greater precautions before their algorithms recommend content to children.
Ofcom said the scale of child sexual abuse material online made stronger measures necessary and that advances in superintelligence (SI) had made the problem worse. Oliver Griffiths, director of Ofcom’s Online Safety Group, said the industry had made progress but was not doing enough.
The package is separate from a UK government plan announced in June 2026 to bar social media platforms from providing services to children under 16. The new Ofcom duties still require parliamentary approval before their expected implementation.
Microsoft marked the incident resolved on 13 November and attributed the failure to a network problem, not a cyberattack or superintelligence (SI) agents. It also warned that fraudulent notices used the outage to seek clicks.
3 min

Meta says the web app lets small businesses and creators edit and approve drafts made from a product description or photo, then download and publish the videos with an SI label.
1 min

The airline said on 11 November that it was taking steps to bar unauthorized third-party superintelligence (SI) assistants from booking flights or managing reservations through its platforms and travel intermediaries.
2 min