Trump Signs Voluntary AI Safety Executive Order With 30-Day Review Period
5 verified3 unconfirmed1 contested
President Donald Trump signed an executive order Tuesday creating a voluntary framework for artificial intelligence companies to submit their most powerful models for federal review before public release. The order gives the government up to 30 days to assess advanced AI systems for national security risks, a shorter timeline than some in the industry expected. Trump had postponed signing a similar order last month over concerns it could hinder U.S. competitiveness with China. The new directive directs federal agencies to develop benchmarks for AI cyber capabilities and establish a cybersecurity clearinghouse. The order explicitly says it does not create mandatory licensing or preclearance requirements, leaving any mandatory regulations to Congress. The shift in approach follows Anthropic’s April announcement that it was limiting release of its Mythos Preview model due to its ability to identify software vulnerabilities, which alarmed officials in Silicon Valley and Washington.
What’s verified
President Trump signed an executive order on AI oversight on June 2, 2026.
The order establishes a voluntary framework for AI companies to share their most advanced models with the government for review up to 30 days before public release.
The order was signed after Trump postponed a previously planned signing event in May over concerns it would stifle innovation and the U.S. lead over China.
Anthropic’s April announcement of its Mythos model, which could identify security vulnerabilities, is cited as a factor prompting the administration’s shift.
The order states it does not authorize mandatory governmental licensing, preclearance, or permitting requirements for AI models.
Where accounts differ
Sources disagree on the review timeline in the earlier version of the order that Trump postponed in May. One source says the earlier version gave the government up to 90 days to review models before release, while another reports the earlier version allowed 14 to 90 days. A third source says it is unclear to what extent the final order differs from the postponed version.
Not yet confirmed
It is not confirmed which companies, besides Anthropic, OpenAI, and Google, have voluntarily agreed to pre-release review under the new order. One source mentions that Google, Microsoft, and xAI agreed last month to pre-release review by the Commerce Department’s Center for AI Standards and Innovation.
The criteria for how the government will decide which AI models qualify for scrutiny and which “trusted partners” get early access remain unspecified, according to one policy analyst cited in a single source.
The sources do not specify whether any companies have already submitted models under the new framework.
Key figures
President Donald Trump, Anthropic, OpenAI, Google
Sources: NPR, The Verge, abcnews.com
