Washington asks OpenAI and Anthropic to stop delivering their models to British testers before US review
According to Politico, the office of the national director of cyber wants to inspect the cutting-edge models before the British security institute accesses them.