The order creates a voluntary framework for AI developers to work with the U.S. government on making their models publicly available, the officials said. Under the framework, developers would be required to provide models to the government 90 days in advance of release, and would also be required to give pre-publication access to critical infrastructure providers such as banks, one of the people said.
Such an approach could be a middle ground among Trump supporters.
MAGA activists, including former Trump adviser Steve Bannon and right-wing political organizer Amy Cramer, are pressuring the White House to require AI developers to submit their most capable models to government security testing.
On the other side of the debate are tech industry advocates such as venture capitalist Marc Andreessen and former Trump adviser David Sachs, who have resisted the mandate. Sachs stepped down from his role as President Trump’s AI chief in March and now serves as co-chair of the President’s Technology Advisory Council. President Trump’s AI policies during his second term largely reflect the perspectives of the technology industry.
Discussion about driving new models
A White House spokesperson called any discussion of the details of AI policy “speculation.” A National Security Agency spokesperson directed Reuters to contact the White House with questions about the details of the president’s plans. National Cyber Commissioner Sean Cairncross, Trump’s chief adviser on cybersecurity policy and strategy, did not respond to requests for comment.
The balance of power between the two groups of Trump supporters has shifted with the release of powerful new AI systems such as Mythos and OpenAI’s GPT-5.5-Cyber. The companies have warned that the new model could intensify complex cyberattacks, but some cybersecurity executives say those concerns are overblown. Mr. Mitos’ arrival sparked a fight among the president’s supporters to influence his response. The outcome of that debate could have a major impact on the AI industry if the president’s decision delays the deployment of large-scale language models or causes companies to change the way their models operate to address safety concerns. Choosing either option can hurt your profits.
Republicans have traditionally favored limited government and opposed regulation, but support for imposing AI guardrails is growing among more vocal populist voters. In a letter sent to the White House last Friday, the populists are calling on President Trump to seek government approval before deploying “potentially dangerous” AI systems.
Kramer said advocating for new regulations is “antithetical” to his political views, but said AI requires a different approach.
“We cannot expect the people leading these AI companies to put our interests first and do the right thing to protect the American people,” she said.
Mr. Kramer helped organize the January 6, 2021, rally that preceded the Capitol riot. She said in an interview that she was not among the thousands of Trump supporters who stormed the Capitol that day.
big tech support
Tech executives are the president’s biggest political donors and among his most visible supporters. At the January 2025 swearing-in, Meta’s Mark Zuckerberg, Amazon’s Jeff Bezos, Google’s Sundar Pichai, and OpenAI’s Sam Altman sat front and center.
Tech industry advocates told Reuters they want the U.S. Department of Commerce’s AI Standards Innovation Center to play a leading role in responding to the Trump administration’s advanced AI models, adding that companies are willing to work with the organization’s scientists and cybersecurity experts on a voluntary basis.
The National Security Agency is participating in administration-wide discussions about how to respond to Mythos, according to Cairncross and two other people familiar with the matter. Lawmakers called on Cairncross to work with federal agencies to establish a process to monitor “sudden leaps in frontier AI capabilities.”
“The past few months have been a massive wake-up call to the kinds of vulnerabilities that AI can create,” said former U.S. Congressman Brad Carson. He currently helps run a super PAC network of which Anthropic is a funder.
Holding off on deploying new AI models while the federal government vets them may give the United States a short-term advantage over adversaries, but it won’t keep the technology out of their hands in the long term, said Neil Chilson, director of AI policy at the Abundance Institute, a nonprofit organization that often works with the tech industry.
“We need to make sure we put it in place and make the most of it, including strengthening our defenses,” Chilson said.
Voluntary federal testing of new AI models has been underway for several years, with companies like OpenAI and Anthropic submitting their products for scrutiny by the Center for AI Standards and Innovation, which was known by another name under former President Joe Biden.
The Commerce Department announced in May that Google, xAI and Microsoft had agreed to submit their AI models to security testing, but the details later disappeared from the website. The White House and the Commerce Department did not respond to requests for comment on why the details disappeared.
