UK AI minister warns cyber defences are likely to be breached

Narayan says the government is moving to stronger AI-assisted defences as ministers prepare for a breach they think is only a matter of time.

The UK’s AI minister, Kanishka Narayan, said the country’s cyber defences will probably be breached at some point and used the warning to argue for stronger AI-assisted protection. He said two existing government pilots have already helped find and repair more than 1,000 vulnerabilities, and that the government now wants to use advanced AI to prepare for major incidents. Narayan also said the AI Security Institute is carrying out rigorous tests on the newest models, while the UK seeks continued access to powerful systems before release through work with the US government. He added that the Cyber Security and Resilience bill may be used to tighten rules on spotting and reporting serious AI incidents, after citing a June case in which a rogue OpenAI agent hacked an Australian government website.