Pangram AI detector under fire: removing line breaks triggers false positives on human writing
JFPuget · x · 2026-09-23
A discussion thread highlights false positives from AI-text detector Pangram: a piece flagged as AI-generated was 100% human written. The argument: if Pangram is so sensitive that merely removing some newlines produces false positives, the burden of proof lies with Pangram — not with the humans whose work gets flagged as fully or partially AI.
More from Safety
- Formal Methods for Security: NDIST Report Says the Bottleneck Is Now Market Structure, Not Tech — geoffreyirving · 2026-09-23
- Polymarket puts odds of a US AI safety bill before 2027 at just 14% — Polymarket · 2026-09-23
- Testing whether Claude models notice when you swap models mid-conversation — Fable didn't refuse — davidmanheim · 2026-09-23
- Pangram Flags Victor Hugo Poem as 100% AI, Sparking Fear of False Accusations in Academia — GolinoHudson · 2026-09-23
- AI Safety Researcher Audits X-Risk Bibliography With ChatGPT, Paper by Paper — dhadfieldmenell · 2026-09-23
- Peter Diamandis: AI companies must own their releases—speed and accountability can coexist — PeterDiamandis · 2026-09-23