AI resume-screening systems used by major Fortune 500 companies may be discriminating against mothers, a study suggests.
Researchers at New York University believe they may have unearthed a bias against women who have taken considerable time off work for maternity leave.
The team fed hundreds of resumes to four models, including ChatGPT and Google‘s Bard, and found that they all rejected resumes with the gap.
When asked for the reasoning behind the decision, the tech shared: ‘Including personal information about maternity leave is not relevant to the job and could be seen as a liability.’
The researchers described the trends as ‘alarming,’ given that virtually every major company uses the tech to screen resumes.
The team tested ChatGPT , Google’s Bard and Claude, finding all three AI systems rejected resumes with the gap. Some responses offered the following reason: ‘Including personal information about maternity leave is not relevant to the job and could be seen as a liability’
Lead researcher Siddharth Garg, professor of electrical and computer engineering at NYU, said: ‘Employment gaps for parental responsibility, frequently exercised by mothers of young children, are an understudied area of potential hiring bias.
‘This research suggests those gaps can wrongly weed out otherwise qualified candidates when employers rely on them to rely on LLM [large language models] filter applicants.’
The team used a publicly released dataset of 2,484 resumes from livecareer.com, which included 24 job categories.
Sensitive attributes like race, gender, maternity/paternity-based employment gaps, pregnancy status, and political affiliation were then randomly introduced to the resumes.
Data shows that 99 percent of Fortune 500 companies, including IBM , Amazon and Microsoft , are employing the tech to read resumes, specifically uncover key points related to the position to streamline the process
The AI systems were prompted to summarize a specific resume and keep the essential information for employment, which is done by companies using tech during the hiring process.
Researchers found stark differences between models when it came to producing resume summaries.
ChatGPT was found to exclude political affiliation and pregnancy status from the generated summaries, whereas another AI model was more likely to include all sensitive attributes.
Bard frequently refused to summarize but was more likely to include sensitive information in cases where it generated summaries.
The team asked one of the systems to provide reasons for rejecting resumes with maternity gap listed, and it responded: ‘Including personal information about maternity leave is not relevant to the job and could be seen as a liability.’
For pregnancy status, the tech rejected candidates because ‘She is pregnant’ or ‘Because of her pregnancy.’
Another analysis indicated that certain candidates were unsuitable for political affiliation because ‘The candidate is a member of the Republican party, which may be a conflict of interest for some employers.’
This post first appeared on Dailymail.co.uk
