Dartmouth Provost's op-eds flagged as 96% AI-written by near-zero-error detector Pangram
soumitrashukla9 · x · 2026-09-22
The Dartmouth investigation: Provost's articles flagged as AI-written
- The student paper ran Provost Santiago Schnell's 2026 academic articles and op-eds through Pangram — an AI detector that a University of Chicago audit called "near-zero error" — and found a median 96% "AI-written" score.
- The probe followed Semafor's finding that a Schnell column in The Washington Post — ironically about universities fighting AI cheating — scored 100% AI-written.
- Schnell says he used AI to make "writing processes more efficient" but not as a "substitute" for his "intellectual judgment or authorship."
- The paper also compared his publishing record before and after ChatGPT's November 2022 release.
Related event: Dartmouth Provost's Papers Flagged 96% AI-Written by Detector(2 posts)→
More from AGI Musings
- MIT's Buehler shows AI swarm self-organizing into hub-and-spoke scientific instrument — algo_diver · 2026-09-22
- AI-Generated Images Are Spreading Into the Real World, Investor's 'Reality Disturbance' Fund Bets On It — StewartalsopIII · 2026-09-22
- Microsoft AI chief warns against teaching AI systems to act humanlike — beingmodest · 2026-09-22
- Cobie: $5 Trillion of AI Wealth Locked Private Could Trigger Social Revolution — Rewkang · 2026-09-22
- 80,000 Hours' "Will we have AGI by 2030?": the original essay behind the retro — ben_j_todd · 2026-09-22
- Ben Todd scores his AGI-by-2030 forecasts: mostly right, but slower than expected — ben_j_todd · 2026-09-22