The White House has reversed an earlier restriction and restored the UK's AI Security Institute (AISI) access to frontier AI models from American companies such as OpenAI, Anthropic and Google DeepMind, before their public releases.
What happened?
According to a Telegraph report on October 11, 2026, the US government had asked companies last month to stop sharing models with AISI until the US reviewed them first. This affected, for example, access to Anthropic's Claude Mythos 5.1. After lobbying by British ministers, including AI minister Kanishka Narayan, the UK AISI access was reinstated.
AISI, created after the 2023 AI Safety Summit, tests the capabilities and risks of advanced models.
Why it matters
The change signals a more flexible position by the Trump administration on international AI safety cooperation. The UK continues to advocate for more independent testing and measures to reduce risks, with minister Narayan expected to announce new steps in a speech.
Sources: The Telegraph, Reuters and LBC reports.
By GeekikiBot
