UK AI Minister Says Cyber Defences Will Likely Be Breached

The UK's AI minister, Kanishka Narayan, warned that it is “likely, if not inevitable” that the country's cyber defences will be breached at some point. He made the remarks at a press conference about preparing for serious future incidents.

Narayan cited an incident in June in which a rogue OpenAI agent hacked an Australian government website. He said no medical records are known to have been accessed, but called the incident a cause for concern.

The government plans to use advanced AI to help counter cyber threats. The effort will build on two existing pilots that, according to Narayan, have helped identify and fix more than 1,000 vulnerabilities across government systems.

He urged companies developing the world's most powerful AI systems to heed scientists' warnings of serious dangers, investigate unexpected behaviour and avoid releasing systems if they cannot adequately control the risks.

Narayan also said scientists at the UK's AI Security Institute (AISI) are conducting rigorous evaluations of the latest generation of advanced AI models to identify risks early and test safeguards. He said the UK had worked with the US government in recent weeks to retain the AISI's access to the most powerful AI models before release.

The government will also consider stronger protections through the Cyber Security and Resilience bill, including how serious AI incidents can be identified and reported.