The White House Takes Control of Frontier AI Access
The White House is now deciding who gets access to frontier AI models, not the labs.
Partner Lists Require Government Approval
Skip to content
Toggle Navigation
News
- Events
- TNW Conference
- June 19 & 20, 2025
- All events
- TNW Conference
- Newsletters
- Partner with us
- Jobs
- Contact
Latest
- Deep tech
- Sustainability
- Ecosystems
- Data and security
- Fintech and ecommerce
- Future of work
- Conference media hub
Startups and technology
- Investors and funding
- Government and policy
- Corporates and innovation
Podcast: Artificial Intelligence
The Trump administration is dictating access to frontier AI models, shifting that decision from Anthropic and OpenAI to the government via the Gold Eagle programme.
Previously, Anthropic controlled access to its Mythos cybersecurity model through Project Glasswing, while OpenAI ran a similar initiative called Daybreak for its cyber model. Going forward, these partner lists will require explicit government approval.
A White House official claimed that the government does not "provide approvals for AI releases" and that company participation is "voluntary." However, recent actions indicate otherwise. The administration blocked Anthropic’s Claude Mythos 5 and Fable 5 last month over national security concerns, reinstating access only after negotiations. OpenAI has also limited new model releases to "trusted partners" due to government requests.
The Gold Eagle programme, launched this week, aims to curate a list of trusted entities for AI vulnerability research, effectively giving the White House authority to approve or deny access to frontier AI models.
This shift is significant as it occurs without legislative action or a dedicated regulatory agency, relying instead on a voluntary programme that has become a de facto gating mechanism.
As David Sacks, former White House AI czar, noted: "This is how you lose the AI race."