Trump Administration Considers Mandatory Pre-Release Vetting of AI Models

May 5, 2026

The Trump administration is discussing an executive order requiring mandatory pre-release vetting of AI models, with the NSA, Office of the National Cyber Director, and Director of National Intelligence potentially overseeing reviews. The proposal resembles the UK's AI Security Institute model and was catalyzed by Anthropic's Mythos model demonstrating autonomous hacking capabilities. David Sacks left his AI czar role in March 2026. A $200 million Pentagon contract with Anthropic collapsed after the company was designated a supply chain risk.

Mandatory AI model vetting proposal

▪The proposed AI model review system would grant the government early access to AI models without blocking the models' release.
▪The NSA, the Office of the National Cyber Director, and the Director of National Intelligence could oversee the AI model review process under the proposed system.
▪The proposed AI model review approach resembles the UK's AI Security Institute model, where government bodies evaluate frontier models against safety benchmarks before and after deployment.
▪The Trump administration is discussing an executive order that would establish a government review process for new AI models before the AI models are released to the public.

White House AI leadership vacuum

▪White House Chief of Staff Susie Wiles and Treasury Secretary Scott Bessent have taken a more active role in shaping AI policy after David Sacks left his role.
▪David Sacks left his role as AI czar in the Trump administration in March 2026.

Anthropic Mythos model controversy

▪The Anthropic Mythos model attracted government attention and appears to have been the catalyst for the Trump administration's AI model vetting discussions.
▪Some studies have found that cheaper AI models can achieve comparable results to Anthropic's Mythos model in vulnerability discovery.
▪Anthropic described the Mythos model as capable of finding thousands of critical software vulnerabilities and too dangerous for public release.

Pentagon contract dispute

▪A $200 million Pentagon contract with Anthropic collapsed.
▪The Pentagon designated Anthropic a supply chain risk after Anthropic refused to remove guardrails on autonomous weapons and mass surveillance.
▪A federal judge called the Pentagon's supply chain risk designation of Anthropic "Orwellian."

Government vulnerability testing access

▪Government agencies other than the NSA remain cut off from Anthropic's tools.
▪The NSA has used Anthropic's Mythos model to assess vulnerabilities in government Microsoft software deployments.