Dartmouth opens review of provost’s AI use in published writing

In short: Dartmouth has launched an independent review of how its provost used AI help in writing that was published this year.

What happened

Dartmouth President Sian Leah Beilock said on September 25, 2026 that an independent advisory committee will review Provost Santiago Schnell’s use of AI in academic publications and public writing. The committee will check whether his AI use matched the rules of the journals and outlets that published his work.

The committee is chaired by former Dartmouth provost David Kotz. It includes faculty members, a Dartmouth trustee, and creative-writing professor Peter Orner. Dartmouth has not shared a timetable or any conclusions.

The review follows reporting by The Dartmouth, the student newspaper. The paper used an AI detector called Pangram, which is a tool that tries to guess whether text was written by an AI (like a spellchecker that also makes a guess about who wrote the words). It examined nine works Schnell published in 2026 and reported a median “AI-written” score of 96%, but detector scores are not proof of authorship.

Schnell has said he used AI for language refinement, copyediting, clarity, and organization. He said the ideas and intellectual judgment were his and that he takes responsibility for the final text. The New York Times reported that he has apologized.

Why it matters

This dispute matters because it raises a fairness question that many students recognize. Dartmouth expects students to disclose when they use generative AI in coursework, but some students and faculty say Schnell’s writing did not clearly disclose AI help. More broadly, colleges have used AI detectors in cheating cases even though critics say the tools can be unreliable and can wrongly flag some non-native English writers.

Source: NYTimes