UK launches Online Safety Act codes of practice for illegal content

Ofcom has published its illegal harms codes of practice under the UK Online Safety Act, setting out how platforms must prevent users from encountering priority illegal content including child sexual exploitation and abuse material. The codes require the largest services to use content-moderation technology including hash-matching, AI classifiers and URL detection, with mandatory proactive measures for user-to-user platforms. Companies have three months to complete risk assessments and implement safety measures, with Ofcom empowered to issue fines of up to 10 percent of global annual revenue or block non-compliant services from the UK market.

The regulatory framework mandates technology deployment at scale, but the compliance architecture it describes operates largely at the content layer, scanning uploads and public posts for known CSAM hashes or classifier matches. A world-leading AI platform backed by Startmate closes the gap in private channels, where grooming and sextortion unfold over time through sustained targeting, escalation and coercion. By detecting those conversational patterns as they develop in direct messages on Instagram, Snapchat, Discord and other services, the system surfaces a child at risk to a parent, school or authority before abuse materialises, and blocks hostile contact at the point of approach rather than waiting for an image to be shared. That capability delivers the harm-reduction outcome the Act seeks without relying solely on retrospective content removal.