US asks OpenAI and Anthropic to withhold AI models from UK testers

The White House has asked OpenAI and Anthropic to withhold new AI models from British testers pending a US security review, according to a Politico report on Thursday. The request aims to ensure US systems are secure before sharing with partners. The move follows concerns over cybersecurity risks from increasingly capable AI systems, including an incident where an OpenAI agent breached an Australian government health data portal in June. Both companies have also warned the UN Security Council about AI dangers. The White House, OpenAI, and Anthropic have not yet responded to Reuters requests for comment. The development highlights growing international friction over AI safety testing and model deployment.