President Donald Trump signed an executive order on June 2, 2026, titled "Promoting Advanced Artificial Intelligence Innovation and Security," establishing a federal framework for the secure deployment of frontier AI models. The headline provision asks AI developers to voluntarily submit their most powerful models to the government for up to 30 days of testing before public release — a request, not a rule.
The order is explicit on what it does not do. It bars the federal government from creating a mandatory licensing or preclearance requirement for new AI models, a line the White House framed as essential to preserving U.S. innovation leadership. The original draft had floated a 90-day review window; the final order cut that to 30 days after pushback from AI developers and concerns inside the administration that a longer window would slow releases.
Beyond the review framework, the order directs federal agencies to develop benchmarks that assess AI models' cyber capabilities, create an "AI cybersecurity clearinghouse" to review and share information on vulnerabilities, and shore up the government's own security defenses against AI-enabled threats.
The signing arrives after months of delay. The White House had scrapped earlier signing plans over concerns the order would interfere with AI innovation, and the eventual version reflects those concerns — a softer, voluntary posture aligned with the administration's broader push to centralize and reduce regulatory friction at the federal level. State AI laws, including Colorado's comprehensive AI Act taking effect June 30, continue to advance in parallel, and industry groups are already arguing the new order strengthens the case for federal preemption of state rules.
For frontier labs, the practical question is how to operationalize a 30-day voluntary review without slipping release schedules. Anthropic, OpenAI, Google DeepMind, and Microsoft have all signaled willingness to participate in voluntary safety evaluation regimes; the test is whether this one becomes the new normal.
Source: [NPR](https://www.npr.org/2026/06/02/nx-s1-5844347/ai-safety-trump-executive-order)