AI Detector Flags Anthropic's S-1 Filing as AI-Written

A user ran Anthropic's 335-page S-1 filing through AI detector Pangram, which flagged large sections of the risk factors chapter, including text on mechanistic interpretability, as AI-generated.

2026-10-27 ~ 2026-10-27 · 2 related posts

• Anthropic's S-1 Risk Section Was Written by AI, Detection Tool Suggests — omooretweets · 2026-10-27
• Anthropic's 335-page IPO filing flagged as AI-written by detector Pangram — andrewjdavison · 2026-10-27