Trump Signs Landmark AI Executive Order

President Donald Trump has signed a revised executive order on artificial intelligence, a move that reverses course after he shelved a prior version less than two weeks ago. Senior aides convinced the president the administration could no longer delay creating a regulatory framework for the rapidly advancing technology, according to two officials familiar with the discussions.

The updated directive grants the federal government access to the most advanced AI models 30 days before their public release. That is a significant reduction from the original proposal, which would have required companies to provide a 90-day advance window. Beyond shortening the review period, officials made few substantive changes to the text.

Trump approved the revised order Monday night after a high-level White House meeting, and aides drafted the final language Tuesday morning, the two officials told WIRED. This marks the first major AI regulation directive of Trump’s second term, reflecting mounting concern inside the administration that increasingly powerful systems, including Anthropic’s Claude Mythos and OpenAI’s GPT-5.5, could be weaponized to launch cyberattacks against critical infrastructure.

The order does not impose formal regulation. Instead, it establishes a voluntary process to identify which AI models are the most powerful. The government then gets exclusive access for 30 days, allowing officials to find and address potential vulnerabilities before the models reach the public.

The move represents a win for White House chief of staff Susie Wiles, who teamed up with Treasury secretary Scott Bessent and National Cyber director Sean Cairncross to revive the proposal, sources told WIRED. They overcame initial resistance from Trump’s former AI czar David Sacks, a prominent skeptic of government intervention in the sector.

With the order in place, Bessent can begin exploring discussions with China about creating a similar cross-border framework for advanced AI systems, according to a person familiar with the matter. Those talks had been paused while the administration finalized its domestic policy, WIRED previously reported.

White House spokesperson Liz Huston said the executive order reflected Trump’s “commonsense approach of collaborating with industry to balance innovation and security, cementing America’s continued global dominance in AI and cybersecurity.”

Several large AI companies signaled support on Tuesday. “This Executive Order is an important step in strengthening America’s leadership in AI. We look forward to collaborating with the White House to support its implementation,” Anthropic wrote in a post on X.

Trump had scrapped an earlier version of the order on May 21 after AI companies and Sacks warned that a 90-day review window would be too burdensome for a fast-moving industry, according to people familiar with the discussions. But executives at several of the biggest AI firms told the administration their models were only becoming more sophisticated and powerful, meaning the White House could not simply postpone a framework indefinitely.

Administration officials worked on the matter through the weekend ahead of Monday’s high-level meeting. Wiles and Bessent attended, along with Sacks, who joined by phone. White House aides involved in the process told some AI companies they expected Trump to eventually sign off on a framework, but they were uncertain about the timeline. In the end, Trump felt there was enough industry buy-in for a 30-day window and gave his approval Monday night.

The order directs multiple federal agencies to create a classified process for determining which AI models the government would want to access, as well as select other “trusted partners” that can also get early access to those frontier models. Beyond the early-access framework, the order instructs the Pentagon to shore up its classified networks within 30 days and directs the Justice Department to bring criminal cases against individuals who use AI models to hack computer systems.