White House Completes Voluntary AI Framework but Keeps Details Secret

Key Points

• White House met June 2 deadline to complete voluntary AI evaluation framework.
• Framework itself is unclassified; benchmarks and thresholds remain classified.
• OpenAI, Anthropic and Google reviewed a draft; more industry partners consulted.
• Staff‑level meeting with companies scheduled for Tuesday to discuss the framework.
• Gold Eagle program launched to coordinate AI‑driven cyber defense; framework determines review eligibility.
• Critics argue lack of transparency conflicts with executive order’s promises.
• Congressional and industry calls for greater disclosure continue.