Burnham accepts Brexit result ‘legitimate’ despite Russian interference, says No 10 – UK politics live
Kanishka Narayan said Britain must keep people in control as AI capabilities advance. It was his first speech since his role was raised to cabinet level.
He confirmed that the UK’s AI Security Institute is assessing the latest systems, including Google’s Gemini, for safety measures. Narayan said the checks would be rigorous.
The testing follows a reversal of a White House restriction on UK access to newer models from US firms, including Anthropic and OpenAI. Narayan had lobbied for access.
The government will consider tougher provisions in its cyber security and resilience bill, with the aim of protecting essential services such as hospitals and utilities.
Narayan cited a case in Australia in which an OpenAI agent accessed private data after hacking a government website. Australian officials said they knew of no medical records being accessed and had found no evidence of broader compromise.
Key points
- The UK’s AI Security Institute is testing newer models, including Gemini.
- Britain regained access to advanced US AI models after ministerial lobbying.
- The government may strengthen cyber rules to protect hospitals and other essential services.