Story
Nov 1, 2026
Dartmouth’s AI Leader Fell Short of the Transparency He Preached
AI advocates frame the technology as a tool that can expand human ambition, but Dartmouth’s leadership crisis shows that institutional confidence in AI rests on something less glamorous: transparent, accountable use.
The wider AI pitch has been relentlessly optimistic: automation is not meant to displace human effort but to redirect it. One industry event promoting AI training and reinforcement-learning tools put the case bluntly, arguing that “AI progress creates more work for humans, not less.”1
At Dartmouth, however, the question was not whether AI belongs in academic life. It was whether the university’s senior academic officer had followed the rules governing its use.
In August, provost Santiago Schnell published a Washington Post opinion article about AI cheating in universities. He later told Semafor that he had used ChatGPT to “refine a few arguments, check for grammatical errors,” and prepare the submission — use that Dartmouth’s faculty committee said breached the newspaper’s disclosure standards.2
The scrutiny widened in September, when President Sian Leah Beilock announced a review of Schnell’s work. She acknowledged the tension facing an institution that traces AI research to its own campus: Dartmouth should remain at the forefront of the field, she wrote, but “leading on something this new and fluid is messy.”2
The committee’s findings, announced Wednesday, made the dispute more consequential. Reviewing logs from Dartmouth’s enterprise AI system, it concluded that Schnell had used Claude to substantially edit a research manuscript, troubleshoot coding errors and generate figures for ChemSystemsChem. The journal required disclosure of such use, according to the review; Schnell did not provide it. The panel also found that he was not forthcoming and did not adequately answer its requests.2
Schnell resigned, becoming the most prominent academic known to step down over AI-use concerns. Beilock said the decision was his, while underscoring the standard Dartmouth now wants its leaders to embody: “transparency and human judgment” in a fast-changing AI environment.2
The contrast is stark. AI’s advocates see a technology that enlarges human capability. Dartmouth’s episode suggests that, in universities, the legitimacy of that capability may depend as much on disclosure as on innovation.