Inside Dartmouth's AI detector policy: why Pangram passed with near-zero error rate
lmoroney · x · 2026-10-12
Full details of Dartmouth's Sept. 29 guidelines: interim dean John Carey authorized faculty to run de-identified student work through Pangram, an LLM-based detector that a University of Chicago audit found to have a 'near-zero error rate'. Rising faculty queries about AI use prompted the school-wide policy.
Related event: Dartmouth Authorizes Pangram AI Detector for Faculty(2 posts)→
More from Safety
- AI recommendations are not approvals: 3 controls in FinSentry payment-agent demo — Proper-Spread-35 · 2026-10-12
- Yandex's Vladimir data center halted after UAV attack, cloud in emergency mode — AccBalanced · 2026-10-12
- Meta Muse contacts permission warns agent "may take unexpected actions. Monitor it carefully." — Heavy-Juice51 · 2026-10-12
- Microsoft President Brad Smith warns AI is 'less regulated than milk' — Polymarket · 2026-10-12
- Can AI liability regulation skip the courts? A two-year lawsuit is an eternity in AI — NathanpmYoung · 2026-10-12
- Security researcher reviews Meta Muse privacy design: 'even more encrypted' isn't a security standard — niloofar_mire · 2026-10-12