The White House has not publicly released details of its voluntary advanced AI model testing framework more than two weeks after introducing it to AI companies. OpenAI, Anthropic, and Google remain unclear about testing standards and government information access.