On May 7, 2026, a Federal Judge in the U.S. District Court for the Southern District of New York ordered the National Endowment for the Humanities (NEH) to reinstate more than 1,400 grants that had been terminated following an Artificial Intelligence (AI)-assisted review process.  The Court also prohibited the government from reallocating the more than $100 million associated with the terminated grants.  See American Council of Learned Societies v. National Endowment for the Humanities, et. al., 1:25-cv-03657 (S.D.N.Y.) and The Authors Guild v. National Endowment for the Humanities, et. al., 1:25-cv-03923, (S.D.N.Y.). 

Court Criticizes Use of AI in Grant Review Process

According to the Court’s decision, individuals affiliated with the Department of Government Efficiency (DOGE) used AI tools to identify grants as “DEI-related” based on terms such as “history,” “culture,” and “identity.” The Court found that grants identified through this process were subsequently terminated because of their perceived viewpoint or subject matter. 

The Court characterized grant terminations as a “textbook example of unconstitutional viewpoint discrimination” and concluded that the government could not rely on AI-assisted processes to justify conduct that violated constitutional protections. Moreover, the Court found that DOGE lacked any statutory authority to terminate NEH grants and had exceeded authority delegated by Congress to NEH itself.

Jurisdictional Issue Addressed by the Court  

The government argued that the claims should have been brought in the U.S. Court of Federal Claims (COFC) because the dispute related to grant agreements. The District Court rejected that argument, finding that the plaintiffs’ claims were grounded in alleged constitutional violations rather than contractual rights.

Relying on recent Supreme Court precedent, the Court explained that while contract damages claims generally belong in the COFC, federal district courts retain jurisdiction over constitutional challenges and claims contesting agency policies or governmental action.

Potential Broader Implications 

The decision is notable for its criticism of DOGE’s use of AI-assisted tools in the grant review and termination process. The Court found that “ChatGPT was the Government’s chosen instrument for purposes of this project, and DOGE’s use of AI to identify DEI-related material neither excuses presumptively unconstitutional conduct nor gives the Government carte blanche to engage in it.” 

In doing so, the Court established an important precedent that the government’s use of AI does not excuse unconstitutional conduct. The ruling may provide a significant framework for challenges in other cases where federal agencies have used similar technology to identify and terminate grants.

Feldesman attorneys continue to monitor litigation developments and administrative actions affecting federal grantees. If you have questions regarding this decision or how evolving federal grants enforcement activities may affect your organization’s federal funding, please contact Phillip Escoriaza, Adam Oppenheim, or Mindy Pava.