UK AI minister calls for stronger state AI capability alongside regulation

Britain needs the ability to build and use AI, not just regulate it, UK AI minister Kanishka Narayan has argued. In a BBC interview, he called for stronger state capability and democratic control, citing a government security exercise that he said uncovered 500 previously unknown vulnerabilities for £13,000 in AI use.

The security exercise behind the argument

Speaking to Nick Robinson on the BBC’s Political Thinking podcast, Narayan described the government using frontier AI models—the most advanced systems—to check for cybersecurity problems in Whitehall. The task he described was finding previously unknown weaknesses in public-sector systems.

Narayan put the spending at £13,000 and said the exercise identified 500 novel vulnerabilities in significant areas of the public sector. He did not disclose what those vulnerabilities were. The figures are the minister’s account of the exercise, rather than a detailed technical assessment that explains the findings.

He also pointed to the UK AI Security Institute as evidence that Britain has a role beyond accepting technology developed elsewhere. Rejecting the idea that only the US and China matter in AI, he said every US lab had approached the institute to request model testing before deployment.

The fundamental thing is that the British public and democracy has to assert power rather than leave it to tech bros and oligarchs

Kanishka Narayan, speaking on the BBC’s Political Thinking podcast

Domestic development as a source of leverage

Narayan’s prescription was to combine regulation with building capability and capacity inside the state. He did not identify particular companies or technology executives as the targets of his criticism. His argument instead concerned who should have the power to shape the technology and its effects.

Building AI in Britain would help the country shape working conditions rather than leave development entirely abroad, he said. He also linked domestic technology to defence, growth and international leverage, arguing that the UK is currently wholly reliant on other countries for essential parts of its technology.

Avoiding another painful transition

The social case for that intervention rested on his comparison with the Industrial Revolution. Narayan described AI’s scale as comparable to, and in some ways greater than, that transformation. He said Britain needed to manage the transition better this time, rather than focus only on the eventual national outcome.

In his account, industrialisation ultimately benefited the country, but the change was extremely painful for too many people. That distinction underpinned his call for public power over AI: a positive result for Britain as a whole would not, by itself, mean the transition had been handled well for those affected.

Narayan also hinted at a possible AI announcement in the following week’s Budget, citing conversations with Chancellor John Healey. He supplied no details, leaving the substance of any Budget measure unresolved rather than setting out a specific new commitment.