Dartmouth Provost Santiago Schnell Resigns Over Undisclosed AI Use
Dartmouth provost Santiago Schnell resigned after an inquiry found undisclosed AI use in an academic paper and a Washington Post op-ed despite disclosure rules.
Summary
By October 30, 2026, Dartmouth provost Santiago Schnell had resigned after an advisory committee found he failed to disclose significant AI use in an academic paper and his August 9 Washington Post op-ed, “Universities Are Fighting AI Cheating. But There’s a Deeper Problem,” despite publisher rules. The committee cited “insufficient cooperation”: Schnell supplied his Dartmouth laptop but not personal LLM logs, and declined an interview. Dartmouth’s Claude enterprise logs showed him requesting referee-style reviews for a ChemSystemsChem submission, code debugging, TeX and repository corrections, and edited or new figures. He also asked whether Academic Freedom and Catholic Mission showed AI-writing signatures, then revised it along Claude’s suggestions. Schnell described AI as an editorial aid for clarity and grammar and said he misunderstood disclosure requirements, but the committee said his public account may understate use whose full scope remains unknown.
In September 2026, Schnell’s “Who Wrote This?” openly detailed GPT-5.6 Pro generating early structures, titles, literature and draft portions; GPT-5.6 Sol editing rewritten sections; and GPT-5.6 Pro plus Anthropic Claude Fable 5.1 reviewing the manuscript. He argued tools can contribute without becoming co-authors if the human checks claims, makes final decisions and accepts responsibility, yet other required disclosures were omitted. Dartmouth, which says AI was launched there in 1956, is proceeding with its “What Must Remain Human?” conference, an Anthropic-sponsored student hackathon and speakers Mira Murati and Yann LeCun. President Sian Leah Beilock, author of September’s “Why I Want More AI at Dartmouth,” said leaders must model transparency and judgment. Survey data found 61% say AI explains difficult concepts, 50% believe AI discussions improve their work, 70% call take-home work unfair, 39% feel peer pressure to use AI, and 92% of Dartmouth students fear damage to writing skills; student Eli Moyse called for Schnell’s firing in September.
Positives
- Schnell’s September 2026 “Who Wrote This?” detailed how GPT-5.6 Pro, GPT-5.6 Sol and Claude Fable 5.1 contributed to the manuscript.
- 61% of surveyed college students said AI helps explain difficult concepts.
- 50% of surveyed students believed discussing issues with AI improves their work.
- Dartmouth’s upcoming conference includes an Anthropic-sponsored hackathon and speakers Mira Murati and Yann LeCun.
- Sian Leah Beilock said Dartmouth leaders must model the transparency and human judgment expected from students and scholars.
Risks & concerns
- Schnell omitted required AI disclosures from an academic paper and his August 9, 2026 Washington Post op-ed.
- Schnell withheld personal LLM logs and declined to meet the committee, prompting a finding of insufficient cooperation.
- Claude logs documented manuscript reviews, code fixes, repository changes, figure work and a request to identify AI-writing signatures.
- 70% of surveyed college students considered take-home work unfair because of AI, while 39% felt pressured by peers to use it.
- 92% of Dartmouth students feared AI could damage writing skills.
- The committee could not determine Schnell’s full AI use and warned that his public descriptions may have understated it.