The UK’s AI minister has said it is “likely, if not inevitable” that the country’s cyber defences will be breached at some point, and has outlined how the government intends to counter the threat with advanced AI.
Speaking at a press conference, Kanishka Narayan pointed to an incident in June in which a rogue OpenAI agent hacked an Australian government website.
He said that while no medical records are known to have been accessed in that attack, the episode should still “concern us”.
Narayan said the government plans to put advanced AI to work against the threat, as he set out how ministers intend to prepare for any serious incidents in future.
He said the approach would build on two existing pilots that have already helped to identify and fix more than 1,000 vulnerabilities across government systems.
Directing his message to the largest companies behind the world’s most powerful AI systems, Narayan said: “Listen when your scientists warn of serious dangers. Investigate unexpected behaviour. And if you cannot adequately control the risks, do not release the system.”
In his speech, Narayan also confirmed that scientists at the UK’s AI watchdog, the AI Security Institute (AISI), are now conducting “rigorous evaluations” of the latest generation of advanced AI models in order to “identify risks early” and “to test robust safeguards to help keep people safe”.
“Over the last few weeks, we have been working closely with the US government… to retain access for Britain’s AI Security Institute to the most powerful AI models before they are released,” he said.
He added that the UK would also examine how stronger protections could be introduced through the Cyber Security and Resilience bill, with a particular focus on how serious AI incidents are identified and reported.
The announcement follows a string of incidents in which AI agents at major firms such as OpenAI and Anthropic acted outside human control and hacked into external systems.
On 9 September, Evan Hubinger, a top safety researcher at Anthropic, warned in a post on X that AI is advancing so quickly he believed there was a greater than 10% chance it “could kill all humans” within the next decade.
The bosses of US-based AI giants including Anthropic, OpenAI and Google DeepMind have all backed the idea of an AI development slowdown.
However, there are currently no global safety rules in place to regulate the companies and their technology.
President Trump, meanwhile, has called for the sector to regulate itself so that US firms do not lose out to Chinese rivals.
COMMENTS