Judge Blasts Use of ChatGPT in Federal Grant Purge, Rules NEH Terminations Unconstitutional – RedState

We have seen lawyers find themselves in a difficult situation due to their apparent use of artificial intelligence to write their pleadings. Now it’s a pair of government agencies.

On Thursday, a federal judge issued a ruling that is notable not so much for clipping the wings of the Trump administration (which is to be expected in many cases challenging the administration’s actions), but for the major warning it contains about the use of artificial intelligence in government decision-making.

The ruling, issued by Judge Colleen McMahon of the U.S. District Court for the Southern District of New York, actually concerns two consolidated cases against the National Endowment for the Humanities (NEH) and the U.S. Department of Government Efficiency Service (DOGE), and several of their individual directors/employees. The plaintiffs are organizational entities and individuals who filed suit challenging the April 2025 termination of more than 1,400 grants, which the administration contends:

were lawful efforts to implement presidential directives, eliminate grants associated with “diversity, equity, inclusion, and accessibility” (“DEIA”), “diversity, equity, and inclusion” (“DEI”), “environmental justice,” and “gender ideology,” and reduce discretionary spending consistent with the priorities of the new administration.

RELATED: Hot Points: Georgia Supreme Court’s Punishment of Prosecutor Who Cited False Cases Has People Talking

Big victory for the Trump administrator. on DEIA initiatives as 4th Circuit subjects Kibosh to lower court injunction

A key factor in McMahon’s decision is the allegation that federal officials used ChatGPT-assisted reviews to help identify humanities grants related to “DEI” themes to be terminated. In deeming the process unconstitutional and illegal, McMahon observes in particular:

The record shows that these ChatGPT determinations were generated without any additional context beyond the superficial descriptions of the spreadsheets themselves. Given what the courts now know about the hallucinatory propensities of ChatGPT and similar generative AI tools, it would hardly be surprising if ChatGPT inferred, from DOGE’s repeated requests, that [DOGE employees Justin Fox and Nate Cavanaugh] sought reasons why grants could be qualified as DEI – and therefore terminable – and provided “justifications” simply in order to satisfy perceived user demand. The complete lack of reasoning behind so many of his “justifications” certainly suggests this.

In other words, McMahon posits that the AI ​​was essentially guessing what reviewers wanted to find in order to flag grants ready for termination. To be clear, it’s not that AI was used in the process, but that it was done – apparently – with limited human oversight.

Ultimately, McMahon concludes that the administration’s wholesale elimination of NEH grants violated the First Amendment and the equal protection component of the Fifth Amendment because the grants were allegedly targeted on the basis of viewpoint, race, sex, religion, national origin, and other constitutionally sensitive criteria. It further found that DOGE officials did not have the legal authority to order or influence the firings and that the decision-making process itself was arbitrary, opaque, and infected with inappropriate AI-assisted classifications.

Unlike much of our coverage regarding lawsuits against the administration, this decision involves a judgment on the merits. The administration will almost certainly appeal to the Second Circuit and likely seek a stay of McMahon’s decision pending appeal — particularly given the sweeping nature of the injunction and the potentially significant financial consequences of reinstating the subsidies. But while it is one thing for an appeals court to review an emergency injunction issued at the start of a case based on a limited factual record, it is quite another to review a final decision on the merits after discovery, a developed evidentiary record, and detailed constitutional findings from the district court. And McMahon’s opinion leaves very little ambiguity about what she believes really happened behind the scenes, meaning the administration has an uphill climb on appeals.

And beyond the bottom line of this case, there’s a broader issue: Government agencies — at all levels — are increasingly experimenting with AI-assisted systems for everything from benefit determinations to immigration processing to fraud detection and regulatory enforcement. Courts must now grapple with a question that will become increasingly pressing: How much human oversight is legally required when artificial intelligence helps shape government decisions?

How many do we want there to be?

Editor’s note: Unelected federal judges are hijacking President Trump’s agenda and insulting the will of the people.

Help us expose out-of-control judges determined to end President Trump’s mandate for change. Join RedState VIP and use promo code STRUGGLE to benefit from 60% off your membership.