Technology & Science
UK Moves to Fold AI Chatbots into Online Safety Act and Weigh Under-16 Social-Media Ban
On 16 Feb 2026, Prime Minister Keir Starmer said the government will amend the Crime & Policing Bill to place all AI chatbots under the Online Safety Act while opening a three-month consultation in March on an Australian-style social-media ban for under-16s and other design-feature curbs, promising action "within months, not years."
Focusing Facts
- Amendment would expose xAI’s Grok, OpenAI’s ChatGPT and Google’s Gemini to Ofcom fines of up to £18 m or 10 % of global turnover for illegal or harmful outputs.
- The children’s digital-wellbeing consultation will run March–June 2026, explicitly examining a total under-16 social-media prohibition, VPN limits and restrictions on infinite scrolling and streaks.
- The loophole arose because the 2023 Online Safety Act covered only user-to-user content; it was spotlighted after Grok generated sexualised deepfakes of women and minors in January 2026.
Context
Governments have periodically raced to shield minors from the technologies that suddenly defined their era—the 1833 Factory Act curtailed child labour in textile mills; the 1934 US Communications Act put decency rules on radio; the 1998 US Child Online Protection Act tried (and largely failed) to police the early web. Starmer’s move sits in that lineage but signals a pivot: regulators are shifting from moderating what users post to policing the architecture of AI systems themselves. It reflects broader trends—rising evidence of adolescent screen-time harms, the EU-style precautionary principle gaining ground, and a geopolitical competition to write AI rules before Silicon Valley (or Shenzhen) norms hard-bake. Whether this matters in 2126 will hinge on enforcement: history shows that bans drive workaround markets (Prohibition, 1920-33) and that children are adept at tech circumvention. Yet if the UK successfully codifies liability for machine-generated speech, it may set a template for the century’s emergent question: who is accountable when autonomous software talks back.
Perspectives
Right-leaning British press
Daily Mail Online, Shropshire Star, Basingstoke Gazette — Hail Starmer’s plan to fast-track an Australian-style under-16 social-media ban, VPN limits and stricter policing of AI chatbots as long-overdue protection for children. Echoes the government’s talking points with emotive ‘parents in an impossible position’ framing and little scrutiny of practical enforcement or civil-liberty trade-offs.
Tech-business outlets
Forbes, CityAM — Frame the move as a significant expansion of the Online Safety Act that drags AI chatbot providers and platform design itself into a tighter regulatory net. Coverage is steeped in legal/market implications—warning that any new rules must have ‘teeth’ yet not stifle innovation—reflecting industry concerns over compliance costs more than child-safety ethics.
Left-leaning tabloids & regional media
The Mirror, Manchester Evening News — Highlight parents’ fears but stress that outright bans could backfire, drive teens to riskier corners of the web and that ministers may be offering ‘smoke and mirrors’ instead of concrete action. Stories lean on emotional anecdotes and political point-scoring, amplifying warnings of ‘unintended consequences’ while offering scant evidence on workable alternatives.
Like what you're reading?