DOGE Staffers Used ChatGPT to Cut Holocaust History Grants During Counter-DEI Purges: Lawsuit

The US Department of Government Efficiency (DOGE) relied on the ChatGPT large language model program when deciding to cut grants for Jewish-related history programs, including one focused on violence against women during the Holocaust, according to a new class-action lawsuit.

DOGE staffer Justin Fox is named as one of the defendants in the suit filed in US federal court on Friday by the Authors Guild, which alleges that he was the one who developed the method of using ChatGPT prompts to determine which grants from the National Endowment for the Humanities (NEH) — also a defendant — to cut in the name of eliminating any programs related to Diversity, Equity, and Inclusion (DEI) initiatives.

Fox said in a deposition that he regarded any grant related to a minority group as qualifying as “DEI” and thus up for elimination. When asked about a grant he chose to cancel related to violence against women during the Holocaust, he responded, “It’s a Jewish — specifically focused on Jewish culture and amplifying the marginalized voices of the females in that culture.” Fox added, “It’s inherently related to DEI for that reason.”

The Trump administration has made a point of targeting DEI programs, especially on university campuses, arguing they foster bigotry by replacing merit with identity-based preferences. Many Jewish groups have criticized DEI initiatives for often excluding Jews, ignoring antisemitism, or characterizing Jews as white “oppressors” rather than as a historically oppressed minority group.

The lawsuit alleges that Fox and his fellow DOGE staffer Nathan Cavanaugh “made and executed the termination decision without any legal authority conferred by Congress. There is no jurisdictional barrier to vacating these unlawful terminations, and permanent relief is warranted.”

One of the projects targeted by DOGE was a translation project titled In the Shadow of the Holocaust: Short Fiction by Jewish Writers From the Soviet Union, which the lawsuit describes as “a critical, annotated translation into English of Yiddish and Russian works written in the aftermath of the most significant Jewish tragedy of the 20th century.”

ChatGPT put the book on the chopping block, stating that “this anthology explores Jewish writers’ engagement with the Holocaust in the USSR.”

According to the suit, the DOGE cuts “are unconstitutional several times over. The record establishes, without genuine factual dispute, that the terminations violated the First Amendment by targeting grants for their viewpoints and perceived political associations; that they violated the equal protection guarantee by classifying grants based on race, sex, and other constitutionally protected characteristics.”

DOGE also allegedly targeted Catholic efforts to promote Holocaust studies. The suit notes that another grant Fox and Cavanaugh chopped was support for the National Catholic Center for Holocaust Education at Seton Hill University.

In a deposition, NEH’s acting chair Michael McDonald said he did not know DOGE had relied on ChatGPT and rejected including Holocaust-related grants under DEI. He also claimed DOGE ignored his disagreements. DOGE possessed the final say about which projects to cut.

In response to the question “In your view, does this grant relate to DEI?” McDonald answered “no.” When the lawyer followed up with “would you consider this to be wasteful spend?” he replied “I would not, no.”

Since the NEH’s founding in 1965, the agency has provided over $6 billion in grants to fund over 70,000 projects in all 50 states.

The lawsuit details that Fox and Cavanaugh lacked “any relevant background in the humanities, public or private grant administration, peer review, or government service of any kind prior to joining the administration.”

According to the filing, the two DOGE staffers met with McDonald and Assistant Chair for Programs Adam Wolfson on March 12. However, Fox and Cavanaugh “entirely controlled the process of selecting grants to terminate and executing the terminations — their approach was top-down, viewpoint- and race-based, and indifferent to the views of NEH leadership or the ordinary processes of grant administration.”

DOGE’s mastermind, billionaire Elon Musk, has a professional rivalry with Sam Altman, the CEO of OpenAI, which developed ChatGPT.

Musk faced multiple controversies last year involving alleged antisemitism, Nazis, and the Holocaust. Following his decision to make a gesture at a Jan. 20 rally which many interpreted as resembling a “Sieg Heil”-style salute, the Anti-Defamation League (ADL) initially defended the billionaire before criticizing his choice to promote Holocaust humor on his X social media platform.

“We’ve said it hundreds of times before and we will say it again: the Holocaust was a singularly evil event, and it is inappropriate and offensive to make light of it. Elon Musk, the Holocaust is not a joke,” Jonathan Greenblatt, the ADL’s CEO and National Director, wrote on X in response to Musk.

Musk faced criticism days later when addressing a gathering of Germany’s far-right Alternative for Germany (AfD) party, where he said, “There is too much focus on past guilt, and we need to move beyond that.” Critics decried the comments, arguing they minimized or dismissed the Holocaust.

“The remembrance and acknowledgement of the dark past of the country and its people should be central in shaping the German society,” Dani Dayan, chairman of Yad Vashem, Israel’s national memorial to the Holocaust, said in response to Musk on X. He warned that not focusing on learning lessons from the past is “an insult to the victims of Nazism and a clear danger to the democratic future of Germany.”

In July, following an upgrade to Musk’s ChatGPT rival Grok, the program began promoting antisemitic conspiracy theories and in one instance labeled itself “MechaHitler.”

Two months later, the Center for Countering Digital Hate (CCDH) and the Jewish Council for Public Affairs (JCPA) released a report revealing the rise of antisemitism on X.

The report utilized OpenAI’s since-discontinued GPT-4o model and drew on a year-long investigation to find that “679,000 posts sampled violate X’s policies on antisemitism, and posts identified as antisemitic got 193 million views in the 11 months of this report, despite X’s promises to limit their visibility. Also, antisemitic conspiracies appear to perform disproportionately well on X, constituting 59% of posts in the sample but 73% of likes.”

The report noted that X had allowed for the rise of so-called “antisemitic influencers” and that “approximately one third of all likes on antisemitic posts were on posts shared by just 10 antisemitism ‘influencers.’ 9 out of these 10 ‘antisemitism influencers’ have more followers on X than any other platform, and 6 out of the 10 are verified on X. 3 out of the 10 profit from paid subscriptions on X.”

Amid the criticism, Musk has denied accusations of antisemitism and said his priority is to make X a bastion of free speech. He visited Israel in late 2023, weeks after Hamas’s Oct. 7 invasion of the Jewish state, and Auschwitz in January 2024. Following the latter trip, Musk said he was “frankly naive” about antisemitism and described himself as “Jewish by association.”

The Tesla CEO and X owner vowed to wear around his neck a dog tag reading “Bring Them Home” that was given to him by a parent of one of the Israeli hostages held in Gaza until all the captives were returned home.