Chatoo Study: Women Docked, Men Credited For Same AI Résumé

The story worth sitting with from Michelle Travis's Forbes column isn't the headline reaction, it is how cleanly the experiment isolates the bias. Zehra Chatoo, the founder of Code For Good Now and a former Meta strategist, built one AI-supported résumé for a marketing position and put it in front of 1,000 UK adults in April 2026. Half saw the applicant as 'Emily Clarke', half saw 'James Clark'. Same résumé, same disclosure that AI helped write it. Only the name changed.

Evaluators who thought the résumé came from a woman were twice as likely to question the candidate's competency, and 22% more likely to doubt her trustworthiness. When the identical work carried a man's name, evaluators were twice as likely to read it as initiative. The framing Travis lands on is that a woman's use of AI was read as inability, while a man's use of the same tool was read as pragmatic problem solving.

The result that hit hardest for me was generational. Gen Z men, the cohort we usually assume has the smoothest relationship with AI, described 'Emily's' résumé as 'weak' 3.5 times more often than the identical 'James' version, which drew a 97% approval rating. That is not a boomer-manager problem being aged out. It is showing up in the youngest cohort in the sample.

Travis's argument is that women being 25% less likely to reach for AI tools than men is not really a skills, access, or interest gap. It is a rational response to a competency penalty in the workplace: if visible AI use gets you docked for competence, opting out is the safer play. Allwork reached the same read on the Chatoo data.

The honest caveat is the shape of the study. It is one experimental setup, one country, one marketing job, one pair of first names, and it tests perception by 1,000 evaluators rather than live hiring decisions. What the reporting doesn't tell you is whether the penalty survives if the AI-use disclosure is removed, or how it interacts with race, seniority, or industry. The forward-looking read is that HR and DEI teams now have a specific, quantified bias vector to point at when their companies push AI-adoption metrics into performance reviews, and the bias-audit vendors have a fresh problem to sell against.