The Trump Administration Is at War Alone on AI Regulations

The Trump administration is facing internal wrangling as officials try to determine whether to revive the AI ​​law that President Donald Trump abruptly ended last month, according to several people familiar with the matter.

Negotiations in the weeks since have been seen as controversial, by major Silicon Valley players and government officials. Some AI executives told WIRED privately that they weren’t sure what the regulatory agency would require, or if anyone would end up signing on.

On May 21, Trump canceled the scheduled signing ceremony just hours before it took place. He told reporters at the time that it could stifle domestic competition and reduce the U.S.’s advantage over China in the AI ​​race.

The most difficult part of the nixed administration was a law creating a voluntary process in which AI labs such as OpenAI, Anthropic, and Google would give the White House access to AI models before they are released to the public to assess their cybersecurity capabilities.

At its core, the push for legislation reflects a recognition within the White House that AI is becoming a national security problem, thanks to the capabilities of Anthropic’s Mythos models and OpenAI’s GPT-5.5, which success in finding problems in legacy software systems. The effort reflects a change in governance.

The law also states that AI labs can provide samples up to 90 days before human release, though several AI executives tell WIRED their companies may not be ready to share samples in advance. Some AI leaders and advocates are hopeful that the regulatory framework can be reformed, with some of the issues less controversial.

Whether the administration can revive the AI ​​administration now depends on the ability of White House officials to cooperate with competing groups, according to aides at several organizations involved in the project.

White House chief of staff Susie Wiles will take charge of a group of officials seeking to install the administration, which also includes Treasury Secretary Scott Bessent and cyber chief Sean Cairncross, a former Republican political aide, aides say.

Bessent has emerged as a recognized force in the AI ​​process. In recent weeks, he met with Anthropic CEO Dario Amodei and other AI managers to forge a path forward, supporters say. And he is expected to take a leading role in negotiating cross-border AI regulations with China.

David Sacks, the former chief of AI, criticizes Wiles. Mr. Sacks told Mr. Trump that his order will be difficult and he says it is difficult well requested to stop signing before the set hours. Politics it has already been said and the Sacks-Wiles dynamic.

In a post on X last week, Sacks wrote, “President Trump understands that bad laws are the driving force behind change in America. Winning the AI ​​race means not only beating China but removing the government barriers thrown up by federal laws and awakening politics in DC.”

But perhaps the biggest obstacle to getting legislation back on the table is Trump himself, aides said. “Resolving these disputes is only as important as getting Mr. Trump to agree,” said a senior administration official who spoke on condition of anonymity to discuss the difficult negotiations.

In a statement, White House spokeswoman Liz Huston said the administration has been trying to figure out how best to implement AI regulations. “The President’s team is united in boldly pursuing his goals and maintaining this complex system,” says Huston.