Ivy provost quit over AI writing; his denial was AI-written too

Dartmouth provost Santiago Schnell resigned on October 28 after a faculty committee found he used AI to meaningfully alter his writing without the disclosure his publishers required — and was not forthcoming about it when asked.

The saga began in August, when Semafor flagged a Schnell Washington Post column as AI-written; Dartmouth's student paper then found nine of his 2026 pieces scored a median 96 percent AI-written by the detector Pangram — and the emailed statement he sent the student reporters scored 100 percent too.

The committee worked from Dartmouth's enterprise Claude logs and emails rather than detector scores, and found he had used Claude to substantially edit a research manuscript and generate figures for a journal that required disclosure. Facing a faculty no-confidence vote, Schnell resigned as provost and left the university.