UK AI Minister Warns of Regulating Advanced Models If Voluntary Testing Fails

ShakeelHashim · x · 2026-08-04

Reuters reported that UK AI Minister Kanishka Narayan stated Britain could regulate advanced AI models if voluntary pre-deployment testing through the AI Safety Institute (AISI) stops adequately protecting the public.

It is important to note that this is framed as a conditional fallback rather than a specific pledge or immediate commitment. In a separate interview, Narayan also discussed broader topics including UK AI chip startups, departmental rebranding, open vs. closed models, and the EU AI Act.

Original post →

More from Safety

Safety channel →