Pangram AI detector flags human-written paragraphs, sparking author fears
ArkhamDreamerZero · reddit · 2026-10-18
After a major French novel was flagged 100% AI by Pangram, an author ran their own tests and found troubling false positives.
- Opening paragraphs of their human-written chapter were flagged as partly AI-generated.
- Testing the full chapter returned 100% human — including the previously flagged paragraphs.
The author asks whether Pangram only works reliably on long samples, and whether short samples systematically produce false positives — a real risk for human writers being wrongly accused.
More from Models
- DeepSeek V4.1-Flash-1017 claimed to hit Opus 5-level coding scores at 83x lower cost — Yuchenj_UW · 2026-10-18
- Claude cracks 26-year-old NSA cipher in 30 minutes, reigniting Bitcoin's secp256k1 debate — jamestagg · 2026-10-18
- Fingerprint test claims DeepSeek V4.1 Flash is actually GPT-6 Astra (unverified) — wateriscoding · 2026-10-18
- UGI leaderboard: Gemini 4 Argon writes best at low thinking, least censored at high — Eissa_Cozorav · 2026-10-18
- Claude cracks a 25-year-old NSA cryptography mystery with $36,864 bounty in under 30 minutes — jradoff · 2026-10-18
- Three Years of AI Progress: 50-Msg GPT-4 Cap to 1M Context, 39% to 96% GPQA — alliekmiller · 2026-10-18