The nonpartisan nonprofit Protect Democracy has filed a lawsuit against four federal agencies—the Office of the National Cyber Director, the Office of Science and Technology Policy, the Treasury Department, and the Commerce Department—to compel the disclosure of a secret framework used by the Trump administration to conduct safety reviews of frontier AI models. The lawsuit alleges that the administration is utilizing a voluntary review process and a clearinghouse known as GOLD EAGLE to decide which AI models are approved for public release, but has withheld the procedural architecture, participant identities, and legal authority behind these decisions from Congress and the public. Protect Democracy argues that this lack of transparency prevents oversight and could facilitate corruption or political retaliation, noting that OpenAI has already negotiated a private agreement to limit model distribution to government-vetted partners. The litigation seeks an order for the release of all non-classified information, including the framework's text and contractual terms, by September 30. The dispute is further complicated by the upcoming congressional debate regarding the reauthorization of the Cybersecurity Information Sharing Act (CISA) of 2015, which may serve as the primary legal basis for the GOLD EAGLE program's information-sharing protections.