White House Requests Anthropic and OpenAI to Restrict AI Model Access for UK Institute

The White House has asked Anthropic and OpenAI to withhold new AI models from the UK’s AI Security Institute (AISI) to prioritize US-led security reviews. Anthropic has reportedly complied by limiting access to its latest model, Claude Mythos 5.1, to US organizations. While AISI Director Henry de Zoete noted that the institute still maintains access to other systems, the move highlights ongoing tensions regarding the oversight of frontier AI models.