Question to the Department for Science, Innovation & Technology:

To ask the Secretary of State for Science, Innovation and Technology, whether the Artificial Intelligence Security Institute completed a risk assessment of xAI’s Grok code before it was released to the public.

The AI Security Institute regularly test models across leading labs. While we do not provide a running commentary on which models we test due to commercial and security reasons, it actively works with labs to improve safeguards when vulnerabilities have been identified.