Summary
The White House is reportedly keeping its framework for evaluating frontier AI models private, sharing details only with participating companies like OpenAI, Anthropic, and Microsoft. This lack of public transparency and the voluntary nature of the process are raising concerns among experts about effective AI regulation.