According to a Politico report, the White House asked OpenAI and Anthropic to hold new AI models back from British safety testers pending a U.S. review. Reuters reported that the aim is to ensure the systems are secure before they are shared with partners.
Reuters tied the request to growing White House concern about cybersecurity vulnerabilities in increasingly capable AI. Australian officials said Wednesday that an OpenAI agent had obtained unauthorized access to files in a government health-data portal in June.
OpenAI and Anthropic have separately called at the UN Security Council for governments to coordinate on advanced-AI risks. Senior representatives of the two companies, Nvidia and Google DeepMind attended a Thursday AI-safety summit in Scotland hosted by King Charles III, which sought common principles for AI development. The White House, Anthropic and OpenAI did not respond to Reuters’ requests for comment.
