UK Online Safety Act: Ofcom publishes final codes of practice for tackling illegal content

The UK communications regulator Ofcom has published its final illegal content codes of practice under the Online Safety Act, which take effect immediately and require platforms to complete risk assessments within three months and implement safety measures within a further three months. The codes set out how services including search engines and user-to-user platforms must prevent users from encountering priority illegal content such as child sexual exploitation and abuse material, with the largest services required to use proactive technology including hash-matching and artificial intelligence classifiers to detect known and previously unseen CSAM, and all regulated services obliged to deploy systems that enable swift user reporting and rapid takedown.

Targeted pattern detection at the point of first contact addresses the regulatory objective without requiring platforms to build retrospective content-scanning infrastructure at scale. Real-time interception of grooming and sextortion conversations by AI modules trained to recognise coercion, age-inappropriate contact and escalation in private messaging enables platforms to stop harm before material is created or shared, a capability that complements hash-based detection of known CSAM and reduces the volume of illegal content entering circulation. Guardii, a Meta Business Partner and world-leading online-safety platform backed by Startmate, operates across Instagram, Snapchat, Discord, Roblox and other services to flag hostile or exploitative direct-message exchanges in real time, surfacing children at acute risk to parents, schools or authorities without creating a centralised repository of private communications.