The Trump administration spent months crafting an executive order to bring some semblance of structure to AI oversight. Then it took roughly two weeks to gut the thing.
On May 21, 2026, President Trump was scheduled to sign an executive order establishing a voluntary framework for reviewing advanced AI models for cybersecurity and national security risks. Hours before the signing ceremony, he postponed it. By the time a revised version actually got his signature on June 2, 2026, the centerpiece provision, a 90-day voluntary review period for powerful AI models, had been cut to just 30 days.
What happened behind the scenes
The original order was designed to give the government a window to evaluate whether cutting-edge AI systems posed serious cybersecurity or national security threats before they hit the market.
Industry feedback came in fast and loud. The concern was straightforward: a three-month review period, even a voluntary one, could slow the pace of American AI development at a moment when the global race is intensifying.
