Question to the Cabinet Office:

To ask His Majesty's Government what steps they are taking to ensure that increasingly capable frontier artificial intelligence models are subject to independent safety evaluation before deployment.

The UK's AI Security Institute (AISI) has built trusted, voluntary relationships with the world's leading AI labs. This collaborative approach lets us keep pace with a fast-moving technology and adapt as it evolves, and has allowed AISI to secure privileged access that has identified dozens of vulnerabilities that developers fixed before public release. AISI is in daily conversation with industry partners and works closely with allies and partners around the world on the risks posed by increasingly capable AI systems.