UK calls on AI firms to boost safeguards, backs sovereign infrastructure
The UK is set to strengthen its cyber defenses, and AI Minister Kanishka Narayan is expected to use a speech in west London to call on frontier AI developers to take more concrete steps to secure their systems, according to The Guardian. Narayan is expected to say the UK must not rely entirely on tech giants to "self-regulate" their models and that the country needs to build more domestically controlled AI infrastructure. The report's backdrop includes the UK AI Security Institute (AISI), which has tested more than 30 models with OpenAI, Google DeepMind and Anthropic and whose findings have led to real changes before deployment, as well as a £500 million sovereign AI venture fund whose first equity investment went to AI infrastructure firm Callosum, plus an AI Hardware Plan and a Cyber Growth Action Plan.
It shows governments stepping up involvement in AI security and domestic compute, which could push frontier labs into more pre-deployment safety changes.