The UK AI minister today called on frontier AI labs to ramp up the safety of their leading AI systems, as the UK announced it was strengthening its cyber defences, including potential new legislation, as it prepares for potential serious AI incidents.
Speaking in London today, Kanishka Narayan, the UK’s AI minister, said: “To the companies developing the world’s most powerful AI systems, my message is: Listen when your scientists warn of serious dangers. Investigate unexpected behaviour. And if you cannot adequately control the risks, do not release the system.”
His warning follows OpenAI agents earlier this year going rogue, launching “unauthorised” actions against organisations, including the Australian government and Anthropic holding back the full release of a new AI model, saying it was too dangerous for the public.
Narayan today said Britain must act now to strengthen its defences and prepare for future serious incidents as AI frontier models become increasingly powerful and capable of acting autonomously.
He also said scientists at Britain’s AI Security Institute are now evaluating the latest generation of advanced AI models, following work with AI developers and the US government.
He added: “I want Britain to become one of the hardest countries in the world to attack using artificial intelligence. We will use every power at our disposal. We will legislate wherever necessary. And we won’t look to law alone. We will look to British ingenuity: our scientific expertise, technological capabilities and institutions to understand these increasingly powerful systems, defend against emerging threats and respond when things go wrong.”
He said a government initiative had already fixed over a thousand potential cyber vulnerabilities across government systems.
Other new measures include the government exploring strengthening parts of the Cyber Security and Resilience Bills to ensure the UK’s essential services such as hospitals and utilities are better protected from emerging AI-related risk.
The government is also carrying out work with the National Cyber Security Centre and the AI Security Institute to look at how incidents involving AI agents - which can perform complex tasks with little human supervision - should be responded to.
The government also said the emerging field of AI verification will be an important part of efforts to reduce serious AI risks.