President Trump on Tuesday signed an executive order establishing a voluntary framework for government review of frontier artificial intelligence models before they are released to the public, ending weeks of internal conflict in the White House over how aggressively to regulate the technology. ordinance titledPromoting Advanced Artificial Intelligence Innovation and Security,” privately signed without the usual live broadcast or public ceremony, contrasting with the fanfare that usually accompanies presidential AI announcements.
The final version is significantly narrower than the draft Trump rejected when he canceled a planned May 21 signing ceremony.It could undermine America’s dominance in artificial intelligence technology.” The original draft proposed a mandatory pre-release review period of 90 days and would give the government formal assessment authority over border models. The signed version asks companies to voluntarily submit models 30 days before launch and participate in a collaborative effort instead of submitting them to mandatory testing.
What does the order do?
The executive order defines three main mechanisms. First, a voluntary pre-release review framework in which AI developers can engage the government to determine whether models under development are appropriate.comprehensive boundary models,” provide access up to 30 days before a planned release, and cooperate in selecting “trusted partners” for early access.The framework is clearly voluntary, meaning companies can opt out without penalty.
Second, within 30 days, the order establishes an AI cybersecurity clearinghouse to be coordinated by the Treasury Secretary, the National Cyber Director, the NSA, and CISA. The clearinghouse will scan for software vulnerabilities, validate discoveries, coordinate remediation and patch distribution, A direct response to the Mythos crisis It demonstrated how AI-detected vulnerabilities can outperform existing disclosure and patching processes.
Third, federal agencies are directed to develop benchmarks for assessing the cybersecurity capabilities of AI models and strengthen the government’s own security defenses against AI-powered threats. The order also applies to security research related to artificial intelligencealthough the specific provisions are less prescriptive than in the original draft.
What was cut
The differences between the canceled bill and the signed executive order represent a victory for the pro-industry faction in the White House. The 90-day mandatory review has been reduced to a 30-day voluntary window. The official government evaluation body was replaced by a collaborative framework. Reporting requirements for companies developing powerful models mirroring provisions in Biden’s repealed artificial intelligence order were eased to avoid what industry allies characterized as regulatory overreach.
Silicon Valley objections to the original project was decisive. AI companies argued that mandatory pre-release testing would slow American innovation, put them at a competitive disadvantage to Chinese firms that face no equivalent requirements, and set a precedent for government oversight of the technology’s deployment. By making participation in the signed ordinance voluntary and establishing a cooperative rather than regulatory role for government, these concerns are addressed.
The void he left behind
The voluntary framework means that the effectiveness of the order depends entirely on whether AI companies choose to participate. Companies that have already pre-launch tested CAISI, including Google, Microsoft and xAI, may continue or expand this collaboration. Companies that find government review commercially untenable or competing to ship products can simply opt out.
The EU’s AI Act comes into full force in Augustcreates a stark contrast: mandatory requirements, legislative powers and penalties for non-compliance. Trump’s executive order sets norms and creates institutional infrastructure (cybersecurity clearinghouse, benchmark development process), but relies on goodwill rather than commitment.
For the White House, a silent signing could be the point. The order gives the administration a policy document it can refer to when asked about AI oversight, creates structures that can be strengthened later, and avoids a public confrontation with the AI industry, whose leaders are among the administration’s most visible supporters. Whether a voluntary framework is adequate for a technology that can detect 10,000 zero-day vulnerabilities a month is a question the order intentionally leaves unanswered.
