President Donald Trump decided on Thursday to indefinitely postpone the signing of an executive order on artificial-intelligence oversight, even as the technology-company executives he had invited to attend the White House signing ceremony were already travelling to Washington. The decision, announced to reporters on the South Lawn shortly after Marine One returned from Camp David, surprised both the assembled executives and his own staff.

The proposed order would have created a voluntary framework under which frontier AI laboratories would share new models with a designated federal testing body for a 90-day pre-release evaluation. CEOs invited to the signing included Elon Musk, Alphabet’s Sundar Pichai, Amazon’s Andy Jassy, Meta’s Mark Zuckerberg, Microsoft’s Satya Nadella, OpenAI’s Sam Altman, Anthropic’s Dario Amodei, and Apple’s Tim Cook.

Asked why he had pulled the order, Trump told reporters: “I didn’t like certain aspects of it. 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.” Tech-industry sources said a sustained lobbying push over the previous 48 hours, including direct phone calls to the president from several of the invited CEOs, had succeeded in convincing him to defer.

Inside the administration the postponement is being read as a victory for the deregulation-first wing of Trump’s tech advisers, led by the White House AI and crypto policy advisor David Sacks, and a setback for officials around national security adviser Mike Waltz, who had pushed for a structured pre-release testing regime as a way of giving US intelligence visibility into model capabilities.

The text of the abandoned order, leaked to Axios over the weekend, would have empowered the National Institute of Standards and Technology (NIST) to run the pre-release tests and would have applied to any model whose training run exceeded 10^26 floating-point operations. Several of the invited laboratories had publicly accepted the framework in the weeks before the signing, including Anthropic and OpenAI; Meta and X/xAI had pushed back hardest.

Reaction split the MAGA coalition. Steve Bannon’s war room called the postponement “a surrender to oligarchy”; conservative AI-safety advocate Tucker Carlson echoed the line. Senate Republican leader John Thune called for a legislative path instead, and at least two Republican senators — Josh Hawley of Missouri and Tom Cotton of Arkansas — said they would introduce their own version of the framework if the White House did not move.

For the AI labs themselves, the postponement removes the most immediate political pressure but does not settle the underlying question. Anthropic CEO Dario Amodei said in a brief statement that his company “remains willing to engage on any reasonable pre-release framework, voluntary or otherwise.” OpenAI declined to comment.