AI news · UK · policy
UK AI minister presses frontier labs to raise safety efforts
Kanishka Narayan’s speech pairs tougher cyber defences with a call for Britain not to rely entirely on labs to police their own models.
Kanishka Narayan, the UK’s AI minister, is set to tell companies building frontier AI that they must do more to keep it safe, in a speech in west London where he is set to announce stronger cyber defences, The Guardian reported on Monday 12 October 2026. In the report, Narayan says Britain cannot rely entirely on the labs to police their own models, and wants Britain to own more of its compute.
AI became a cabinet-level brief under Prime Minister Andy Burnham, and this is the first speech on it since. The same report says MI5, MI6 and GCHQ chiefs briefed the cabinet on AI threats last week, according to the Sunday Times, and that Geoffrey Irving, once chief scientist at the UK AI Security Institute, fears hostile states could take models before release.
Narayan has set out the line before. At the Labour conference he told Fortune that testing models is “good but clearly insufficient” and that countries must harden their defences, The Next Web reports. He named cyber attacks and AI help with biochemical weapons as his main worries, and said the UK wants to shape global AI principles during its G20 presidency next year.
Britain is moving from testing labs’ models towards demanding more of the labs themselves, and that pressure reaches every firm shipping AI into the UK.
Sources
- The Guardian: UK to strengthen cyber defences as minister calls for AI firms to increase safety efforts. Checked
- The Next Web: UK AI minister says AI testing is not enough, Fortune reports. Checked
Alexandre S.
Alexandre S. is a creative ad strategist and copywriter with ten years' experience in digital marketing across multiple channels.