SACRAMENTO, the United States, May 27 (Xinhua) -- The largest empirical study to date on the use of AI in hiring found "clear racial disparities in applicant outcomes," according to a Stanford-led study published Tuesday.
Researchers from Stanford University, Chapman University and Northeastern University said AI is reshaping not only whether firms hire, but also how they hire. More than 90 percent of U.S. employers now use hiring algorithms to screen applicants, they noted. The study analyzed data from 3.4 million applicants who submitted 4 million job applications to 156 employers across 11 sectors.
The study, titled "Algorithmic Monocultures in Hiring," found that 14.74 percent of applications submitted by Asian candidates and 25.87 percent of those submitted by Black candidates were directed to positions where the screening systems had an adverse impact on those groups under U.S. employment discrimination standards.
"To put this in perspective," the authors wrote, "if the AI had recommended Black and Asian candidates at the same rate as the most-favored group (typically white applicants), 40,000 more of their applications would have advanced to the next stage of the hiring process."
The researchers also examined what they termed "algorithmic monoculture," a phenomenon in which multiple employers rely on the same vendor's hiring algorithms. Under such arrangements, a negative algorithmic assessment can potentially affect a candidate's prospects across several companies using the same system.
The study recommended that employers and auditors assess automated hiring tools at the level of individual job positions rather than relying solely on company-wide or vendor-wide results. The study did not determine whether any employer violated the law or whether rejected applicants would ultimately have been successful hires. ■