The big picture. President Trump postponed plans to sign an executive order on artificial intelligence that would have established a voluntary review process for advanced AI models before public release. The order was effectively shelved by senior tech-industry advisers to the administration.

The order.

  • Would have created a voluntary pre-release review framework for frontier AI models.
  • Federal agencies would have had up to 90 days for security review.
  • OpenAI and Anthropic were reported to be negotiating the framework with the administration.
  • Drafted in response to Anthropic’s disclosure of its Mythos model, which prompted national-security concern inside the executive branch.

Trump’s stated reason. “I didn’t like certain aspects” of the order, the president said. He added that the order “gets in the way,” reiterating that the United States is “leading China, we’re leading everybody” and that he did not want to do anything to compromise that lead.

Why it matters. Reporting requirements for foundation-model trainers — what the federal pre-release review would have institutionalized — have been the central federal question for the labs since EO 14110 was revoked in January 2025. The May postponement leaves the question open, but with the administration’s stated position now clearly tilted against pre-release review.

Who pushed back. Senior tech-industry advisers to the White House, plus a faction inside the administration that takes the “do not slow U.S. AI leadership” posture as load-bearing.

Who wanted it. U.S. intelligence agencies and national-security officials, who, per the Washington Post, have been seeking more sway in AI regulation.

What’s next. No successor signing timeline has been announced. Industry attention turns to whether the administration revises the order, lets it lapse, or routes the function through agency rulemaking instead.

Sources

Sources