Trump Halts AI Order Over Fears It Could Hurt US Edge Over China

In brief

• Trump said he postponed an AI executive order signing because parts of the draft could have slowed U.S. AI development.
• The proposal would have created a voluntary framework for AI companies to share frontier models with the government before public release.
• Trump also said Chinese President Xi Jinping acknowledged the U.S. lead in AI during a meeting last week.

President Donald Trump said Thursday he halted a planned signing ceremony for his administration’s AI executive order after deciding parts of the proposal could undermine the U.S. position in the global AI race with China.

“I didn't like certain aspects of it,” Trump told reporters in the Oval Office. “I think it gets in the way of—we’re leading China. We’re leading everybody. And I don’t want to do anything that’s going to get in the way of that lead.”

Trump did not specify which specific provisions he opposed or say when the administration might revisit the order.

Trump’s comment came during a press conference announcing the administration’s rollback of environmental regulations affecting refrigeration equipment, part of a broader deregulation push the White House says will lower costs for consumers, protect jobs, and reverse Biden-era policies it considers harmful to industry.

First announced earlier this month, the executive order would have established a voluntary AI review framework under which participating companies would provide the federal government with early access to advanced models before public release for national security and capability testing.

The proposal would also have allowed critical infrastructure providers, including banks, to receive pre-release access to covered models, and included a cybersecurity provision aimed at identifying vulnerabilities in unreleased AI models.

The proposed executive order comes as U.S. officials grow increasingly concerned about the national security implications of advanced AI models, particularly after Anthropic’s Claude Mythos demonstrated the ability to identify hundreds of software vulnerabilities and autonomously execute complex cyber operations during testing. At the same time, AI companies including OpenAI, Google, and have expanded with U.S. defense and intelligence agencies, with the on classified networks despite an ongoing with Anthropic, according to .