The intersection of artificial intelligence and federal governance has moved from the realm of theoretical debate into a high-stakes legal battle. New discovery documents from a lawsuit against the National Endowment for the Humanities (NEH) have revealed that the Department of Government Efficiency (DOGE) utilized ChatGPT to identify and terminate humanities grants, often by tagging them as “DEI programs.”
The revelations come via a motion for summary judgment filed in the U.S. District Court for the Southern District of New York on March 6, 2026. The filing, brought by a coalition of prominent academic organizations, suggests a process where funding decisions were outsourced to an AI model and executed by staffers who may have lacked the legal authority to make such determinations.
At the heart of the controversy is the DOGE termination of humanities grants, a move that plaintiffs argue bypassed congressional oversight and ignored the expertise of the NEH’s own professional staff. The discovery process has uncovered a systemic reliance on AI-driven keywords to slash spending, leading to the cancellation of research projects ranging from the study of the Holocaust to postwar Soviet Jewish literature.
The AI Methodology: Tagging “DEI” via ChatGPT
According to court documents, the Department of Government Efficiency employed a flawed process using ChatGPT to flag grants for termination. The AI was tasked with identifying programs associated with “Diversity, Equity, and Inclusion” (DEI), which were then slated for funding cuts as part of a broader effort to reduce federal spending.
The impact of this methodology was particularly acute for projects focused on Jewish history and culture. A class-action lawsuit brought by the Authors Guild and the Association for Jewish Studies alleges that the AI tagged Jewish-themed grants as “DEI.” In one specific instance, a grant focusing on violence against women during the Holocaust was flagged for termination. Justin Fox, a DOGE staffer identified in the suit, testified that ChatGPT properly classified the grant as DEI because it was “specifically focused on Jewish cultures” and the “voices of the females in that culture,” according to the filing detailed by JWeekly.
This AI-driven sweep resulted in the cancellation of approximately 80% of the NEH’s grants. The plaintiffs argue that this process lacked nuance and failed to distinguish between ideological programming and legitimate historical and cultural research.
Legal Breaches and the Ceding of Authority
Beyond the use of AI, the discovery documents highlight significant concerns regarding administrative law and government transparency. Depositions from key figures—including Adam Wolfson, NEH Assistant Chair for Programs, and Michael McDonald, the NEH General Counsel and Acting Chair from March 2025 to January 2026—reveal a breakdown in the traditional chain of command.
The filings allege that Michael McDonald effectively ceded his authority over the grant process to DOGE staff. In a communication to DOGE staffer Justin Fox, McDonald reportedly wrote, “as you’ve made clear, it’s your decision on whether to discontinue funding any of the projects on this list.” This shift in power is a central point of the lawsuit, as plaintiffs contend that DOGE team members made funding decisions despite having no legal authority to do so, effectively cutting out the role of Congress in the funding process.
the lawsuit alleges a violation of the Federal Records Act. Discovery revealed that DOGE and NEH staff utilized the encrypted messaging app Signal to communicate about the grant termination process, a practice that avoids the public record and undermines government accountability, according to a release from PRNewswire.
A Study in Contrasts: The Tikvah Fund
While hundreds of grants were being terminated via AI keywords, the NEH simultaneously awarded its largest-ever grant to a single entity. The agency bestowed $10.4 million to the Tikvah Fund, a politically conservative Jewish cultural project. This stark contrast in funding—where AI was used to purge specific cultural studies while a massive sum was granted to a politically aligned organization—has become a focal point for the plaintiffs’ claims of arbitrary and capricious decision-making.
The lawsuit emphasizes that some grants were terminated even after NEH staff concluded that the projects did not conflict with the Trump Administration’s new policies. This suggests that the AI’s “DEI” tag overrode the professional judgment of the agency’s subject-matter experts.
Who is Affected by the Grant Cancellations?
The fallout from these terminations extends across the United States, affecting academic institutions, libraries, and community organizations in all 50 states and the District of Columbia. The organizations leading the legal charge include:
- The American Council of Learned Societies (ACLS)
- The American Historical Association (AHA)
- The Modern Language Association (MLA)
- The Authors Guild
These groups argue that the reliance on AI to determine the value of humanities research is a fundamental failure of critical thinking—the very skill that humanities education is designed to foster.
What This Means for Federal Governance
The case against the NEH and DOGE serves as a cautionary tale regarding the “automation” of government policy. The core legal question is whether an executive task force can use an unverified AI tool to override established funding commitments and statutory authorities.
If the court grants the motion for summary judgment, it could set a major precedent regarding the legality of using AI for administrative decision-making and the requirement for human oversight in the distribution of federal funds. The plaintiffs are seeking the full restoration of the canceled funding to ensure that critical historical research—particularly that which explores marginalized voices and sensitive historical periods—is not erased by a keyword filter.
| Issue | Alleged Action | Legal/Policy Concern |
|---|---|---|
| Decision Tool | Use of ChatGPT to tag grants as “DEI” | Lack of nuance; arbitrary targeting of Jewish studies |
| Authority | Acting Chair ceded power to DOGE staff | Lack of legal authority to terminate grants |
| Transparency | Use of Signal for official communications | Violation of the Federal Records Act |
| Equity | Canceled 80% of grants while funding Tikvah Fund | Inconsistent application of policy |
As the legal process continues, the case highlights a growing tension between the drive for government “efficiency” and the protection of academic freedom and administrative due process. The outcome will likely determine how—and if—AI can be used to manage the federal budget without violating the law.
The next critical checkpoint in this case will be the court’s ruling on the motion for summary judgment filed on March 6, which will determine if the case can proceed to trial or if the court will rule immediately on the legality of the grant terminations.
Do you believe AI should have a role in determining federal funding for the arts and humanities? Share your thoughts in the comments below or share this article to join the conversation.