AI Models Favor some Accents over Others, Researchers Find
Researchers said accuracy checks are not enough after the models gave lower scores to Chinese and Nigerian accents despite understanding the same scripted speech.
5 Articles
5 Articles
Indian English accent faces hidden bias in AI-powered job interviews and tests, University of Washington study reveals
A University of Washington study has found that leading AI models tend to favour American and British English accents over Indian English when evaluating job interviews, presentations and spoken language tests. Researchers tested four AI models across five English accents and found that Indian speakers received middle-range ratings despite models accurately transcribing their speech. The findings highlight potential accent bias in AI-based recru…
AI models favour US and UK accents, score Chinese and Nigerian ones lowest, study finds
Ask an artificial intelligence model to grade a job interview or test answers, and a person’s accent may change the score. That is according to new University of Washington research that looked at how some of the most advanced AI models rated English speakers with different accents, both on what they said and how they said it. In the scripted tests, every speaker said the same words, and AI models transcribed the accented speech accurately, mean…
A research from the University of Washington found that some advanced AI models differently qualified speakers who read the same scripts: U.S. and UK accents received better grades, while Chinese and Nigerian accents were at lower levels.
Coverage Details
Bias Distribution
- 50% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium







