White House orders voluntary frontier AI reviews

The White House published a new executive order on June 2, 2026, setting up a voluntary federal review process for the most advanced AI models before they are released more widely. The order, titled Promoting Advanced Artificial Intelligence Innovation and Security, gives U.S. agencies 60 days to create a classified benchmarking process for advanced cyber capabilities and to decide when an AI system should count as a covered frontier model.

The most important part is the review window. AI developers would be able to give the federal government access to covered frontier models for up to 30 days before those systems are released to other trusted partners. The order says this access must be governed by confidentiality, cybersecurity, insider-risk and intellectual-property protections. It also says the framework is voluntary and does not create a mandatory licensing, preclearance or permitting system for AI model releases.

That distinction matters. For AI companies such as OpenAI, Anthropic, Google, Microsoft, Meta and xAI, the order signals more federal attention to frontier models without immediately imposing a hard approval regime. For users and businesses, it shows that the policy debate is moving from broad AI principles to specific questions about cybersecurity, critical infrastructure and early access to powerful systems.

The order also directs agencies to strengthen federal cyber defenses and create an AI cybersecurity clearinghouse with industry and critical-infrastructure operators. The White House specifically names organizations such as rural hospitals, community banks and local utilities as examples of systems that could need better access to AI-enabled cybersecurity tools.

The timing is significant because advanced AI models are increasingly discussed not only as productivity tools, but also as systems that can find vulnerabilities, write code and support complex cyber operations. The administration is trying to balance two competing goals: keeping the United States ahead in AI development and making sure the most capable systems are understood before they reach sensitive users or infrastructure.

The result is a lighter-touch policy than a formal regulator, but still a clear marker. Frontier AI is now being treated as part of national security planning, not just as a commercial technology race.