The abstract is read at the publisher; the summary is JClub's.
In audit experiments with randomly assigned applicant gender, LLMs generate gender-stereotypical CVs and show gender bias in screening scores and wage recommendations.
- Why it matters: As generative AI begins mediating hiring tasks like CV writing and applicant evaluation, it may amplify existing labor-market inequalities through biased generation and scoring.
- What they did: The study runs a series of audit experiments where gender is randomized while other CV information is held constant, then measures screening scores and wage recommendations across prompts and LLMs.
- The result: Disparities reflect evaluation bias and generation bias, with screening scores tracking occupational gender segregation and wage recommendations tracking gender wage gaps when gender is explicit, implying measurable algorithmic amplification risks in AI-mediated job search.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.