President Trump signed this executive order on June 2, 2026. The order, titled “Promoting Advanced Artificial Intelligence Innovation and Security,” marks a shift from the administration’s earlier strongly deregulatory stance on AI. It encourages (but does not mandate) tech companies to voluntarily submit their most powerful “frontier” AI models for government review up to 30 days before public release. 1️⃣ Voluntary pre-release review: Companies are asked to give federal agencies (involving national security and cyber experts) access to advanced models for testing, particularly for cybersecurity risks like discovering zero-day vulnerabilities 2️⃣ Shorter timeline: The final version uses a 30-day window (reduced from a proposed 90 days in an earlier draft that was shelved due to industry and competitiveness concerns) 3️⃣ AI Cybersecurity Clearinghouse: The Treasury Secretary is directed to establish a new body to review security vulnerabilities found by AI models and coordinate fixes with agencies and companies 4️⃣ Classified benchmarking: Agencies (including the NSA) will develop processes to assess advanced cyber capabilities of frontier models.