AI text detectors struggle when language models mimic an author's style
Quick Answer
AI text detectors like Pangram, GPTZero, and Originality.ai excel at identifying plain AI-generated text but struggle significantly when models mimic specific authors' styles, particularly in scientific writing, where detection failure rates can reach 29%.
Quick Take
Epoch AI's study reveals that up to 20% of such imitations go undetected, highlighting a critical gap in current detection technologies.
Key Points
- Pangram, GPTZero, and Originality.ai missed up to 20% of style-imitated AI texts.
- Detection failure rates in scientific writing reached 29% for Originality.ai.
- Pangram missed 48% of Gemini-generated academic passages.
- All detectors performed well on plain AI text, with false-negative rates under 1%.
- Epoch AI's study emphasizes the limitations of current AI text detection methods.
Source Excerpt
From the original publisher, up to about 700 charactersEpoch AI tested three leading AI text detectors (Pangram, GPTZero, and Originality. ai) using style-imitated texts. Up to 18 percent of AI-generated passages went undetected. For scientific writing, the miss rate climbed as high as 48 percent, the very genre where these detectors likely see the most real-world use.
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from The Decoder
See more →
An AI model programmed nonstop for 19 days on a single MirrorCode task that cost $2,600 to run
Epoch AI's MirrorCode benchmark reveals Claude Opus 4.7 as the leader with a 56% solve rate, reconstructing a 16,000-line toolkit in 14 hours. Despite this, all models tested struggle with the most complex tasks, highlighting limitations in current AI capabilities. The single task consumed $2,600 over 19 days, raising questions about cost-effectiveness in AI development.
