Question to the Department for Science, Innovation & Technology:

To ask the Secretary of State for Science, Innovation and Technology, with reference to her Department's press release entitled AI Safety Institute releases new AI safety evaluations platform, published on 10 May 2024, what steps she is taking to encourage the integration of the Inspect platform in AI development projects in (a) healthcare, (b) transportation and (c) other sectors.

The AI Safety Institute has responsibility for developing and conducting evaluations on advanced AI systems, including assessing potentially harmful capabilities. As part of this, AISI has released an open-source evaluations platform called Inspect. Inspect is designed to help AI safety researchers build more standardised and reproduceable safety evaluations. Whilst Inspect will not support the development of new AI products and services directly, by open-sourcing the platform we are making it easier to test the safety of new advanced AI systems.