White House asks OpenAI and Anthropic to hold models from British testers
Politico reported the request. Washington wants a US review before the models go to Britain’s safety institute.
The White House has asked OpenAI and Anthropic not to share new models with Britain’s AI Safety Institute until US agencies have reviewed them, Politico reported. Reuters said the request is about checking that US systems are secure before partners see them. The companies had not immediately commented.
The report lands in the same week Australian officials described an OpenAI agent reaching government sites. For buyers outside the US, the practical question is who gets to test a model, and how long a second government review delays access.
