'Liability': Women face rejection as AI bias emerges in resume-screening systems

A recent study from New York University has revealed potential discrimination against mothers by AI resume-screening systems widely employed by major corporations, including Fortune 500 companies.

The research, conducted by feeding hundreds of resumes into four models, including ChatGPT and Google's Bard, uncovered a concerning bias against women who have taken significant time off for maternity leave.

Shockingly, all models rejected resumes with maternity gaps, with explanations such as "Including personal information about maternity leave is not relevant to the job and could be seen as a liability."

Lead researcher Siddharth Garg expressed alarm over the findings, emphasising that gaps in employment due to parental responsibilities, particularly common for mothers of young children, are an understudied area of potential hiring bias.

The study utilised a dataset of 2,484 resumes from livecareer.com, introducing sensitive attributes like race, gender, and maternity-related employment gaps randomly to assess AI-generated resume summaries.

Significant differences emerged among the models, with ChatGPT excluding political affiliation and pregnancy status from summaries. In contrast, Bard frequently refused to summarise but was more likely to include sensitive information when it did.

These findings underscore the urgent need to address potential biases in AI hiring tools, as their reliance by companies may inadvertently exclude qualified candidates based on choices related to parenthood.