AI Study Finds Identical Resumes Rated More Harshly When Attributed to Women
A new study published May 10 in Fortune found that AI-generated resumes containing identical credentials were rated as weaker when evaluators believed the applicant was a woman. The disparity held across multiple reviewer groups.
Women who used AI tools in the hiring process faced harsher judgment than men presenting the same material.
What the Study Found
Researchers created matched resume pairs using AI. The content was word-for-word identical.
Evaluators consistently scored the women’s versions lower on perceived ability and trustworthiness. The Fortune report cited one key finding from the research team: “If people believe they will be judged more harshly for using AI, they are less likely to adopt it, regardless of their capability.”
The study did not limit its findings to hiring managers.
It tested a range of evaluators. The pattern held across groups, suggesting the bias is not confined to one professional context or demographic of reviewer.
Researchers framed the problem as a self-reinforcing loop.
Women who anticipate negative judgment for using AI tools may avoid those tools altogether. That avoidance then widens the skills and productivity gap between men and women in AI-augmented workplaces.
Why This Matters for Workplace AI Adoption
The findings arrive at a moment when AI tools are being embedded across white-collar work.
Employers and productivity software vendors are actively pushing AI-assisted writing, analysis, and communication. If adoption is not uniform across gender lines, the productivity gains from AI will not be distributed evenly.
The study does not suggest AI itself is biased in generating the resumes.
The bias lives in how human evaluators respond to those resumes once gender is inferred. That distinction matters for how organizations respond.
Technical fixes to AI output will not solve a problem rooted in human perception.
No regulatory body has yet addressed AI tool bias in hiring outcomes. The Equal Employment Opportunity Commission has issued guidance on AI screening tools, but the question of differential adoption and social penalty for AI use sits outside existing frameworks.
Background
The concern that AI tools could amplify existing labor market inequalities has grown alongside adoption rates.
A parallel line of research has documented that automated resume screening systems encode historical hiring biases. The new study shifts focus from the screening side to the applicant side, asking whether women face social costs for using AI that men do not.
That framing is newer. The Pentagon’s ongoing effort to integrate AI into its workforce, covered this week, illustrated a broader pattern of uneven AI adoption across different groups.
What Comes Next
Researchers said the findings point to a need for organizations to actively normalize AI tool use across gender lines, not just provide access.
Passive availability does not equate to equal adoption when social penalties are asymmetric.
The study did not measure salary or promotion outcomes linked to AI use gaps. Follow-on research is expected to track whether differential adoption translates into measurable wage divergence over a two-to-three year horizon.
Read Next: Nvidia’s AI Investment Bets Surpass $40 Billion in 2026
