US President Donald Trump said on Thursday that he had postponed the signing of a planned executive order on regulating artificial intelligence (AI), out of concern that it might undermine America’s leading position over China in the field.
“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 told reporters in the Oval Office.
The signing ceremony, which was originally scheduled for Thursday afternoon, was delayed “because I didn’t like certain aspects” of the new regulations, Trump said.
According to earlier reports, the executive order would have required developers of AI models to submit their models to several federal agencies for voluntary review up to 90 days before their public release.
“We have a very substantial standard on AI, it’s causing – it’s causing tremendous good, and it’s also bringing in a lot of jobs, tremendous numbers of jobs,” Trump said, adding, “again, we have more people working right now than we’ve ever had. I really thought that could have been a blocker.”
US media, citing sources, reported that the decision to initiate formal regulatory procedures stemmed from concerns that AI was becoming increasingly powerful and could pose a security risk to the US.
Last month, a new AI model called Mythos, launched by Anthropic, sparked an unprecedented reaction among policymakers and regulators worldwide due to its powerful ability to identify and exploit cybersecurity vulnerabilities.
Anthropic did not publicly release the model but instead provided it to a consortium of US companies, including Cisco, JPMorgan Chase and Nvidia, for use in a programme called “Project Glasswing” to safeguard their critical software.
Previously, in March, Anthropic sued US Defence Secretary Pete Hegseth and the US Department of Defence after the company attempted to restrict the use of its technology in the military sector, leading the Pentagon to classify it as a supply chain risk, a label typically reserved for foreign adversaries.
The planned executive order comes at a time when the AI race between China and the US is intensifying, particularly in the contest over cutting-edge models.
At the summit held in Beijing earlier this month between Chinese President Xi Jinping and Trump, the two leaders agreed to launch an intergovernmental dialogue on AI, according to the Chinese Foreign Ministry.
No clear plan for Trump administration to regulate AI
Meanwhile, the Trump administration has yet to put forward a clear plan to regulate AI, out of reported concerns that any regulation would curb the technology’s development.
In early March, the Trump administration unveiled an AI policy framework aimed at curbing state-level restrictions on innovation, whilst urging Congress to address AI risks and consolidate US leadership amid an increasingly intense technological competition with China.
Congress should pre-empt state AI laws that “impose undue burdens”, to ensure a national standard instead of “fifty discordant ones”, Trump said in a legislative recommendation named the National AI Legislative Framework.
Lizzi C. Lee, a fellow at the Asia Society Policy Institute’s Centre for China Analysis, said the proposed executive order reflects a broader trend in both the US and China, with governments trying to determine where the regulatory frontier should sit for frontier AI – particularly as models become more agentic and cyber-relevant, such as Anthropic’s powerful Mythos.
How Trump’s executive order affects the US-China AI speed race depends “on how heavy the review process becomes”, Lee explained.
“If it stays light-touch and narrowly focused on national security risks, it probably won’t slow leading US labs much. But it does show that Washington is also moving away from a purely laissez-faire model of AI development, which I think is good news for AI security/governance,” she said, adding the development is also relevant to the broader AI governance conversations following the recent Beijing summit between Trump and Xi.
Advanced AI models are not ordinary consumer software
According to Lee, even amid geopolitical competition, both the US and China are converging around the idea that the most advanced models cannot be treated as ordinary consumer software.
“I think a separate, potentially more important race is on governance and safety: not just about who has the most advanced models, etc, but who can govern powerful AI without choking off innovation,” she added.
Kyle Chan, a fellow in the John L. Thornton China Centre at the Brookings Institution in Washington, said that “AI safety and regulation can be done in a way that doesn’t compromise innovation”.
The US has yet to enact a comprehensive law regulating AI, and globally, such legislation remains scarce. China’s regulatory process is accelerating significantly.
A legislative work plan for 2026 issued in May by the State Council, China’s cabinet, outlined plans to “improve AI governance and accelerate comprehensive legislation for the sound development of AI”, using such detailed language for the first time.
At the same time, the National People’s Congress, China’s national legislature, listed AI legislation as an item to be reviewed for the third year in a row.
In early April, Beijing also issued another new regulation requiring Chinese companies engaged in artificial intelligence to establish internal “artificial intelligence ethics review committees”. -- SOUTH CHINA MORNING POST
