Dartmouth provost resigns after review finds undisclosed AI use violated publication rules
The committee examined Claude logs and emails, not AI-detector scores. It found disclosure failures but recommended no disciplinary action.
Loading page…
The committee examined Claude logs and emails, not AI-detector scores. It found disclosure failures but recommended no disciplinary action.
Listen to this story
Dartmouth’s review makes this a governance and disclosure case, not a blanket finding against AI-assisted writing: it found violations in specific work while clearing one coauthored paper. Schnell is leaving after findings that his AI use breached publication requirements and Dartmouth standards; the committee did not recommend discipline, and the president said the departure was his choice. The investigation examined only a handful of at least 30 pieces and lacked requested logs from personal AI accounts, leaving important limits on the record as universities and publishers set disclosure expectations.
The committee reviewed Dartmouth enterprise Claude logs and emails, not AI-detector results, but lacked requested logs from Schnell’s personal accounts.
For one journal manuscript, Claude substantially edited text, troubleshot code, and generated figures; the journal required disclosure, which Schnell did not provide.
The committee called Schnell’s ChatGPT-assisted Washington Post column a clear breach of that newspaper’s policy, not a universal rule for AI editing.
Dartmouth Provost Santiago Schnell has resigned and is leaving the university after a faculty-led review found that his AI-assisted writing violated publication requirements and university standards. President Sian Leah Beilock announced the departure on October 28, 2026, saying senior leaders must model the transparency expected of students and scholars.
Beilock described the resignation as Schnell’s decision, made out of respect for his colleagues and Dartmouth. The committee did not recommend disciplinary action. Those are distinct parts of the outcome: its report set out findings about his conduct, while the president’s announcement said he had chosen to leave.
Beilock formed the committee on September 25 after news coverage questioned Schnell’s AI use, including an August Semafor article that named his Washington Post column. Former provost and computer science professor David Kotz led the review. The committee’s report was dated October 27, the day before Dartmouth announced the resignation.
The inquiry examined records of Schnell’s activity rather than using software to guess whether his prose was machine-generated. According to Valley News, the committee reviewed logs from Dartmouth’s enterprise Claude tool and his emails. It explicitly said it did not use AI-detection tools.
Its access was incomplete. Schnell provided his Dartmouth-issued laptop for inspection, but the committee said he did not supply requested logs from personal AI accounts used for academic and public writing. It also limited the review to a handful of at least 30 pieces he had published since joining Dartmouth in July 2025, citing the short timeline.
For one journal paper, the committee found that Schnell used Claude to substantially edit a research manuscript, troubleshoot errors in computer code and generate figures. The journal required disclosure of that work, but he did not disclose it. The finding concerned several kinds of assistance, not only changes to grammar.
Schnell told Beilock in a September 22 email that he had relied on older author instructions. The committee said the publisher had required AI disclosure since at least 2023. His explanation therefore did not resolve the committee’s concern about whether the journal’s disclosure requirement had applied to the manuscript.
The newspaper example involved ChatGPT rather than Claude. Schnell had told Semafor that he used ChatGPT “to refine a few arguments, check for grammatical errors, and submit it to the Washington Post.” The committee called that a “clear violation” of the Post’s policy. Its conclusion addressed the newspaper’s stated rules, not a universal standard for AI editing.
For a forthcoming article, the committee said Schnell uploaded his manuscript to Claude and asked whether it contained many signatures of AI writing. He later submitted a version with changes of the kind Claude suggested. When the journal editor asked about AI use, Schnell said he had not understood that such editing required disclosure. The committee said that description understated his actual use.
The findings were not uniformly adverse: the committee found no violation in a fourth paper that Schnell co-authored. Its review thus distinguished between individual publications rather than treating all the writing it examined as a single violation.
That is why, as leaders, it is so critical that we model the transparency and human judgment that we expect from our students and scholars.
Dartmouth President Sian Leah Beilock, October 28 announcement
Dartmouth’s announcement also cited the committee’s finding that Schnell was not forthcoming about his AI use. It said his conduct fell short of the university’s Code of Ethical Conduct, Academic Honor Principle and Geisel professionalism policy. The institutional concern extended beyond whether a particular publication had received the disclosure it required.
Schnell did not respond to Valley News’s request for comment on October 28. Dartmouth has named history professor Matthew Delmont interim provost through June 2028.
Loading discussion...
Join the conversation
Explain whether leadership responsibilities should change the consequences.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.