Santiago Schnell, the provost of Dartmouth College, has resigned after a faculty-led review found that he used artificial intelligence to “meaningfully alter” his published writing and was not forthcoming about it. The case began with a guest column in The Washington Post about AI cheating at universities.
Dartmouth’s president, Sian Leah Beilock, announced the resignation to campus on 28 October 2026. “This was his decision,” she wrote. The provost is the university’s second most senior official and its chief academic officer.
How the case began
In August, the news site Semafor reported that Pangram, an AI-detection tool, had flagged Schnell’s Washington Post column of 9 August as AI-generated. The column was titled “Universities are fighting AI cheating. But there’s a deeper problem.”
The student newspaper, The Dartmouth, then ran his other recent publications through the same tool, which flagged most of them as largely AI-written. His writing from before 2022 did not show the same pattern, the Associated Press reported.
Schnell said at the time that he had used ChatGPT to “refine a few arguments” and check grammar, and that the column reflected his own thinking.
What the review found
On 25 September, Beilock set up an independent committee led by former provost David Kotz. Its eight-page report, dated 27 October, did not rely on AI detectors. Instead, it compared his writing with each publisher’s AI rules and reviewed his emails, his university laptop and his logs from Dartmouth’s enterprise version of Claude, an AI assistant. According to Inside Higher Ed and New Hampshire Public Radio, the committee found that:
- In a chemistry journal paper, he used AI to substantially edit the manuscript, troubleshoot code and generate figures, without the disclosure the publisher required.
- For the Washington Post column, his use of ChatGPT was a “clear violation” of the paper’s policy, which requires authors to confirm their work was not created or altered with AI.
- For a forthcoming article, he asked an AI tool whether his draft showed signs of AI writing, then submitted a revision that used its suggestions. The committee said his account to the editor understated his AI use.
- In a fourth paper, co-authored with others, it found no violation.
The committee said his cooperation was “insufficient” because he did not provide logs from his personal AI accounts when asked. It did not recommend disciplinary action.
Schnell’s response
Schnell has acknowledged using AI tools, but said they helped with language and clarity and did not replace his research, judgement or authorship. After resigning, he told The Washington Post that he had not been given an adequate chance to respond to the committee’s findings. He joined Dartmouth in July 2025 from the University of Notre Dame, where he was dean of the College of Science.
Reactions on campus
Criticism had been building for weeks. In September, the Dartmouth chapter of the American Association of University Professors launched a petition calling for his removal. “How do we ask our students and colleagues to follow the rules when the academic leader of Dartmouth is allowed to break them?” it asked, Inside Higher Ed reported.
Students also pointed to a double standard. Dartmouth’s honour code requires students to disclose any use of generative AI in their coursework, and the provost’s office oversees those rules. Inside Higher Ed reported that his resignation came less than two weeks before a planned vote of no confidence by faculty leaders.
Beilock, who has encouraged wider use of AI at Dartmouth, said leaders must “model the transparency and human judgment that we expect from our students and scholars.” History professor Matthew Delmont will serve as interim provost until June 2028.
A wider question for universities
The case has added to a debate across higher education about where acceptable AI assistance ends and dishonesty begins. As Forbes noted, many universities have clear AI rules for students, but fewer for staff and senior leaders. Observers also pointed out that the committee reached its conclusions from records and publishers’ policies, not from AI-detection scores, which can be unreliable on their own.
What students should know
Disclosure is the key issue. The committee’s concern was not that AI was used, but that it was used without the disclosure that journals, publishers and the university required.
Check the rules every time. Universities, departments, journals and scholarship bodies each set their own AI policies, and they change often. Read them before you use AI on an essay, thesis, research paper or application, and keep a record of how you used it.
When in doubt, declare it. A short statement of how you used AI is far easier to explain than a use that is discovered later.
Sources: Associated Press; Inside Higher Ed, 29 September and 28 October 2026; The Dartmouth, 28 October 2026; New Hampshire Public Radio, 29 October 2026; The Washington Post, 28 and 31 October 2026; Forbes, 30 October 2026; Valley News, 28 October 2026. Checked on 1 November 2026.