Style Imitation Breaks AI Text Detectors
The Decoder · rss · 2026-07-19
Epoch AI tested three mainstream AI text detectors (Pangram, GPTZero, Originality.ai) on texts designed to "imitate an author's writing style."
Results show that this type of style imitation significantly weakens detection efficacy: up to 18% of AI-generated paragraphs went undetected; in scientific writing scenarios, the false negative rate reached as high as 48%. The article points out that this is precisely the domain where these detectors are most likely to be deployed in real-world applications.
More from Safety
- Ben Bajarin: Frontier AI Labs Are Becoming the Enterprise Cyber-Defense Stack — BenBajarin · 2026-10-26
- Ben Bajarin: Frontier AI is becoming the enterprise cyber-defense stack — BenBajarin · 2026-10-26
- DetectorScope turns black-box AI text detectors into auditable ones with sparse autoencoders — _akhaliq · 2026-10-26
- Anthropic's IPO filing devotes 94 pages to existential AI risks, from self-preservation to blackmail — Polymarket · 2026-10-26
- Study finds AI reviewing AI poorly detects reward hacking; comprehension audits proposed — ronbodkin · 2026-10-26
- A 13KB Markdown file can hijack Claude Code and Codex via your Downloads folder — pbaylies · 2026-10-26