Dartmouth Provost Schnell Faces AI-Detection Scrutiny Over Pangram Results

Dartmouth’s investigation, using Pangram, found high AI-written percentages in Provost Santiago Schnell’s post-2022 writings (median ~96%), while several pre-2022 works appeared 100% human-written. Schnell says AI was used as an assistive tool and that he authored and reviewed final content, but critics and Pangram proponents note debates over detector accuracy and bias. The case also touches on Schnell’s Washington Post op-ed and related editorials, and how outlets verify submissions’ provenance in the AI era.