President Donald Trump signed a new executive order aimed at tightening U.S. oversight of “frontier” AI models, asking companies to share new models with the government up to 30 days before wider release and pushing agencies to build an “AI cybersecurity clear
On Tuesday, President Donald Trump signed an executive order that puts the U.S. government closer to the most powerful AI systems—at least in how those models move from a developer’s lab to the public.
The order seeks more oversight of “frontier” artificial intelligence models and. for the first time in this administration’s posture toward the technology. signals a clear departure from a more hands-off approach. It asks technology companies to voluntarily share new AI models with the government for up to 30 days before releasing the models more widely.
That shift comes with an additional request: companies should collaborate with the administration to “select trusted partners” that will gain early access to the models. The stated aim is to “promote secure innovation and strengthen the cybersecurity of critical infrastructure.”
The executive order doesn’t stop at model-sharing. It directs leadership of the U.S. Department of the Treasury. as well as the Office of the National Cyber Director. the Department of Defense. the National Security Agency. and the Department of Homeland Security to work with the Cybersecurity and Infrastructure Security Agency to develop an “AI cybersecurity clearinghouse.”.
That clearinghouse is intended to work with the tech industry and with infrastructure operators such as power companies and hospital administrators to identify and fix AI software vulnerabilities.
The order’s push to strengthen U.S. resilience to cyberthreats—and to guard against potential rogue AI actors—marks a major shift from the Trump administration’s previous stance. which had been more laissez-faire compared with the previous Biden administration’s push to make the AI industry more accountable and more geared toward safety. also on a voluntary basis.
It is also striking in its timing. Trump had just recently backed away from a previously proposed executive order on AI safety. That earlier effort was inspired in part by the release of Anthropic’s model Mythos, which the company itself said was too dangerous to be publicly released.
The questions now feel immediate: how much oversight will follow the voluntary sharing request. and how quickly will an “AI cybersecurity clearinghouse” translate into real fixes for systems that run the country—power grids. hospitals. and the organizations that rely on software that may increasingly include AI.
Editor’s Note (6/2/26): This is a developing story and may be updated.
Trump executive order frontier AI models AI cybersecurity clearinghouse Cybersecurity and Infrastructure Security Agency trusted partners critical infrastructure cybersecurity National Cyber Director NSA Department of Homeland Security Department of Defense Department of the Treasury