Trump Signs Scaled-Back Executive Order on Preventing AI Cyber Threats

President Donald Trump signed an executive order Tuesday asking some artificial intelligence companies to give the federal government 30 days to review their frontier models before release.

Trump had been scheduled to sign a broader version of the order on May 21, but he delayed the signing, saying he did not like “certain aspects” of the order. The change of heart reportedly followed calls to the president from former AI czar David Sacks and other tech industry leaders.

The order signed Tuesday does not impose mandatory licensing, pre-clearance, or permitting requirements for new AI models. It calls for a voluntary framework through which AI developers can “engage the Federal Government to determine whether model(s) under development meet the designation of ‘covered frontier model.'”

The framework will also “provide the Federal Government with access to covered frontier models, subject to appropriate confidentiality, cybersecurity, insider-risk, and intellectual-property protection, use, and nondisclosure requirements, for a period of up to 30 days before developers plan to release the models to other trusted partners.”

“Advanced AI capabilities make our Nation stronger, but also introduce new national security considerations that require coordinated action across executive departments and agencies (agencies), and components,” the order says.

The order directs various agencies, including the Committee on National Security Systems, the Department of War, the Department of Homeland Security, the Office of Management and Budget, and the National Cyber Director, to prioritize cybersecurity initiatives.

“As these capabilities evolve, my Administration will continue to work closely with industry to ensure that the best and most secure technology is deployed rapidly to confront any and all threats to our country,” the order continues.

White House officials were motivated to draft the order due to increased awareness of national security risks posed by new models such as Anthropic’s Mythos, which has demonstrated the ability to identify major software vulnerabilities.