The White House has called staff from OpenAI, Google, and Anthropic in Tuesday to review a finished AI oversight framework, according to Leo Schwartz and Stephanie Palazzolo at The Information. The framework gives labs a voluntary route to hand new models to the government for review before any release to partners or the public. The Office of the National Cyber Director is running the meeting, which the companies are sending staff to, not executives. Whether the administration wants feedback or will present a final text isn't clear, though an official said Monday the document met its August 1 deadline. Trump ordered it in early June after Anthropic withheld Mythos over cybersecurity concerns, and has since restricted exports of Fable and made OpenAI stagger GPT-5.6's release.
