Research question
Are AI-text detectors fair across writers?
Key findings
- Seven detectors flagged 61% of TOEFL essays by non-native speakers as AI-generated - Simple prompting reduces false positives and defeats detection
GPT detectors are biased against non-native English writers
The study behind universities abandoning AI detectors, showing systematic bias against non-native English writers.
Contribute
Spot an error, or know a record we should add? Members can propose additions and submit corrections.