On June 2, 2026, the White House issued an executive order titled “Promoting Advanced Artificial Intelligence Innovation and Security.” It frames U.S. policy as advancing leadership in AI while managing the national security risks of increasingly capable systems, and it pursues that on two tracks: hardening government and critical-infrastructure cyber defenses with AI-enabled tools, and building a voluntary framework for the secure development and release of “covered frontier models.”
The order is deliberately voluntary on the frontier-model side. Developers can engage the federal government to determine whether their models qualify as covered frontier models and can give the government time-limited pre-release access, up to 30 days before broader release, under confidentiality, cybersecurity, insider-risk, and intellectual-property protections, and can designate trusted partners for early access. The text states plainly that nothing in it authorizes a mandatory licensing, preclearance, or permitting requirement - a direct contrast with the pre-clearance ideas debated in earlier years. On cyber, it directs CISA to expedite the defense of civilian federal systems and to stand up programs supplying AI-enabled defensive tools, with agency deliverables on aggressive 30- and 60-day timelines, and it directs the Attorney General to prioritize enforcement against AI-enabled cybercrime.
The order matters because it stakes out the federal government’s preferred posture: encourage rapid AI development, keep oversight voluntary and non-licensing, and concentrate hard requirements on cyber defense and criminal enforcement rather than on gating model releases. That stance also sets up tension with states that have moved toward mandatory frontier-model rules, making the federal-versus-state balance one of the defining AI-governance questions of 2026.