Dartmouth’s AI Champion Fell to the Disclosure Rules He Urged Others to Follow
Dartmouth’s AI Champion Fell to the Disclosure Rules He Urged Others to Follow
Santiago Schnell’s fall unfolded against a backdrop of his own warnings about artificial intelligence in higher education. On Aug. 9, 2026, Dartmouth’s provost published a Washington Post essay, “Universities are fighting AI cheating. But there’s a deeper problem.” Later, investigators found he had not disclosed AI assistance used on that article, despite the newspaper’s rules. He had told Semafor he used chatbots “to refine a few arguments” and “check for grammatical errors.”
The faculty advisory committee’s review went further. Using records from Dartmouth’s Claude enterprise account, it found Schnell had used the tool to substantially edit a ChemSystemsChem manuscript, troubleshoot code and generate figures—uses that the journal then required authors to disclose. The committee also said he did not provide sufficient answers to its requests and failed to fully cooperate with the inquiry.
Schnell’s position, as described in his later writing, was not that AI should secretly replace scholars. He argued that authorship and software contribution are different things, and that a human who selects the question, checks sources, revises the argument and makes final decisions remains accountable. Yet the committee said his public explanations may have understated the nature of his actual AI use, leaving a sharp gap between that theory of responsible assistance and his publication practice.
In September, President Sian Leah Beilock acknowledged that Dartmouth’s ambition to remain at the forefront of AI would be “messy,” with missteps part of institutional learning. By Wednesday, Schnell had resigned, making him the most prominent academic known to leave a senior university post over AI-use disclosures. Beilock said the central obligation for leaders was to model “the transparency and human judgment” expected of students and scholars.
Dartmouth will press ahead with AI initiatives, but the Schnell episode has made its immediate lesson unusually plain: using the technology may be defensible; obscuring its role is not.
Write a comment