A new study surveyed the AI company Pymetrics and found that 15% of Asian applicants and 26% of Black applicants applied to jobs where the artificial intelligence tool appeared to be biased against their racial group.
Stanford University researchers analyzed a dataset of more than 4 million applications between 2018 and 2022 for nearly 2,000 positions. The researchers calculated that 40,000 more Asian and Black candidates would have been recommended for jobs to employers if not for the algorithmic racial bias in the AI screening tool.
Significance. As the number of applicants has increased, more and more companies rely on AI hiring tools to lighten the load for hiring managers.
The Stanford Institute for Human-Centered AI found that 90% of U.S.-based companies use AI hiring tools to seek out and evaluate candidates. With AI hiring tools being used by most U.S. companies, these racist practices are bound to affect a large number of Black and Asian candidates on the job hunt.
Overlooked angle. Racist practices in hiring are hardly new. Previous research has found that hiring managers are more likely to exclude job applicants with names normally associated with people of color and those when applicants have braids, locs, and other hairstyles associated with Black people.
The power structure. According to one study conducted on Western countries, racist practices in hiring continue despite laws banning it, suggesting that it could be stemming “from enduring stereotypes, prejudices, or racist ideologies.”
Further attacks on diversity, equity, and inclusion programs under the Trump administration continue to erode antiracist protections in hiring. Recently, Andrea Lucas, the Trump-appointed chair to the Equal Employment Opportunity Commission, wrote a letter to leaders of Fortune 500 companies opposing racial equity. “We are the Equal Employment Opportunity Commission, not the Equitable Employment Outcomes Commission,” Lucas said.
What experts say. “A lot of prior studies had shown racial bias in hiring when people are making the decisions,” said Dan Jurafsky, co-author of the study and a professor of linguistics and computer science at Stanford University. “It was surprising that AI systems that use game-based assessment to rank people were still biased against Black and Asian applicants.”
Deeper dive. The American Civil Liberties Union offers resources to help people understand their rights during the hiring process when they suspect algorithmic racism. Their resources can be found here.
