Distrust in recruiting: 49% of employed job seekers say AI recruiting tools are more biased than humans
However, 39% of Gen Z, Millennials report using AI tools in job hunt
Nearly half of employed U.S. job seekers (49%) believe artificial intelligence (AI) tools used in job recruiting are more biased than their human counterparts, according to the latest American Staffing Association Workforce Monitor® online survey conducted by The Harris Poll.
The news comes just weeks after the Equal Employment Opportunity Commission released guidance on how to incorporate AI into a job search while still adhering to the Civil Rights Act of 1964 and the Americans with Disabilities Act, and amid ongoing discussions of AI screening tools acting with inherent bias when making important decisions.
The skepticism is higher among individuals actively involved in the job-seeking process. Among those who are at least considering a new job, 43% believe AI recruiting tools are more biased than humans, compared to just 29% of those with no immediate plans for a job change.
At the same time, the study found that 39% of current job seekers have used AI tools to assist in applying for a job. Usage varies based on race/ethnicity among U.S. residents, with 36% of Hispanic and 34% of Black U.S. adults saying they have used AI when applying for a job, compared to only 17% of White Americans.
“Job seekers may feel comfortable using artificial intelligence tools in their job search, but that does not equate to trusting AI to make fair hiring decisions,” said ASA chief executive officer Richard Wahlquist. “As AI tools become more widely deployed, it’s critical that hiring managers work to increase transparency and accountability in their hiring processes and use tools that meet current and emerging antibias standards. It’s also critically important that policymakers and technologists thoughtfully consider measures intended to lower bias in AI hiring systems.”
This survey was conducted online within the U.S. by The Harris Poll on behalf of ASA from June 20–22, 2023, among a total of 2,037 U.S. adults age 18 and older, of whom 1,225 were employed. The sampling precision of Harris online polls is measured by using a Bayesian credible interval of +/-2.7%.