AI text detector language bias

Occurred: July 2023

Can you improve this page?
Share your insights with us

Seven AI writing detection tools 'frequently' misclassify non-native English writing as generated by AI systems, according to a study by Stanford University researchers. 

The findings raise questions about the accuracy and reliability of AI writing detection tools in general, as well as about their potential to discriminate against non-native English speaking students, academics, and job applicants.

The researchers ran English essays written by non-native English speakers through seven popular GPT detectors to see how well the AI detection systems performed. 

Over half were classified as AI-generated By contrast, over 90% of essays written by native English-speaking eighth graders in the US were classified as human-generated by the same systems.


Country: USA; Global

Sector: Education; Business/professional services

Purpose: Detect AI writing

Technology: NLP/text analysis; Neural network; Deep learning; Machine learning
Issue: Bias/discrimination - language


Page info
Type: Issue
Published: July 2023