Elon Musk's DOGE Department Used ChatGPT to Classify Museum Funding as DEI

According to court filings, Elon Musk's Department of Government Efficiency (DOGE) utilized ChatGPT to screen federal humanities grants for diversity, equity, and inclusion (DEI) content. The chatbot classified a museum's $349,000 HVAC grant application as DEI-related.

This screening process was detailed in a lawsuit alleging that DOGE violated the First Amendment and equal protection clause by cutting National Endowment for the Humanities (NEH) grants. DOGE, an informal White House advisory body established on Trump's first day in office, operated without formal structure until November 2025.

DOGE staff Justin Fox and Nate Caverno reportedly used ChatGPT to assess whether NEH proposals aligned with DEI criteria. The North Carolina High Point Museum applied for funds to upgrade aging systems for artifact preservation; ChatGPT deemed the upgrade beneficial for more visitors and categorized it as DEI-related, leading to grant cancellation.

DOGE canceled the grant despite only two out of 11 HVAC-related proposals being classified as DEI. Another example involved a Vermont museum proposal addressing climate risks to buildings, which ChatGPT tagged as DEI based on environmental sustainability in heritage preservation.

The department screened 1,163 NEH grant applications using ChatGPT, marking 1,057 for cancellation and retaining only 42. This resulted in approximately half of the agency's annual budget being cut. NEH Acting Chairman Michael McDonald stated many targeted projects were unrelated to DEI expansion, suggesting cost-cutting rather than ideology drove the decisions.

Four academic and literary organizations have filed a lawsuit challenging the cuts, with the case expected to test whether using AI chatbots for DEI classification is legally valid.