Question to the Department for Science, Innovation & Technology:

To ask the Secretary of State for Science, Innovation and Technology, what assessment he has made of Google's compliance with the AI safety commitments made at the Seoul AI Safety Summit in May 2024 in relation to its Gemini 2.5 Pro model.

We expect all signatories to the Seoul commitments to stand by their agreements. The AI Security Institute, within DSIT, has ongoing discussions will all major developers, including Google DeepMind, about the implementation of frontier AI frameworks that guide the safe development of AI.

The government welcomes Google's recently published framework that prioritises the emerging risk of deception in AI models and their plans to publish safety cases.