White House restores UK access to pre-release US AI models
After a policy that demanded US sign-off before firms shared unreleased AI systems with British safety testers was scrapped, the UK AI Security Institute regained early access to new models from several major developers, the UK's AI minister confirmed Monday.
By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .
This story is no longer being updated. It reflects reporting up to .
What happened
The White House reversed a policy that had required US approval before AI companies could share new models with British safety testers. The change restores the UK AI Security Institute's early access to models from Anthropic, Google DeepMind and OpenAI. The UK's AI minister confirmed Monday that Britain will keep early access to the latest models for safety testing.
Why it matters
The UK can again test the latest AI models for safety before they are released to the public.
Timeline
- September: The Trump administration told AI companies to stop sharing their models with the British agency.
- Oct 12, 7:36 AM ET: The White House reversed the policy that had required US approval for sharing new models with UK testers.
- Oct 12, 8:52 AM ET: The UK's AI minister confirmed Britain will retain early access to the latest models for safety testing.
More from Technology & AI
- Eight states sue OpenAI, alleging misleading AI safety claims
- OpenAI announces solution to Navier-Stokes math problem
- Datadog bets AI agents can handle testing of AI-written code
- Anthropic and OpenAI sketch wealth-sharing plans before IPOs
- Mamdani administration proposes AI customer service rules
- Florida engineers study how AI behaves in high-risk situations