A recent study by researchers from the Fox School of Business found that the use of AI avatars in early-stage job interviews could make hiring more objective.
Researchers from the Fox School of Business found that study participants evaluated job candidates based on their qualifications rather than the race of the AI avatar representing them.
Photo courtesy of Alexa Trifilo and created using AI.
First impressions matter, especially in the hiring process, but what happens when an employer’s bias affects their ability to fairly judge a job candidate? Could removing visual cues about a candidate help level the playing field?
Researchers from Temple University tested this concept in a recent study that explored whether job candidates using AI avatars in early-stage interviews could make hiring more objective.
They found that study participants consistently evaluated job candidates based on their qualifications rather than the race of the avatar representing them, suggesting AI avatars could help standardize the early stages of the hiring process.
“The participants were picking up on the cues that we wanted them to pick up,” said Alexa Trifilo, a 2025 graduate of the Fox School of Business’ DBA program who led the research. “They consistently distinguished between stronger and weaker candidates, and the race of the avatar did not affect how candidates were scored.”
“Implicit Bias in Hiring: Leveraging Avatars to Foster More Inclusive Hiring Practices,” was published in Engaged Management ReView on May 18, 2026. The study was led by Trifilo, who now works as a solutions engineering director at Cisco. It was co-authored by Crystal Reeck, associate professor of marketing, and Gary Blau, professor of human resource management, both from Fox.
“As a female in engineering, I’ve always been concerned about diversity in the workforce,” Trifilo said. “I was thinking about what we could do from a hiring perspective to eliminate some of that implicit bias that people may have about women in technology specifically. Can we use AI avatars to hide somebody’s identity in an interview process to make the decision-making more fair?”
For the study, 1,167 participants watched simulated first-round job interviews involving one of three AI-generated avatars: one designed to appear white, one Asian and one intentionally designed to have no clearly identifiable race. The researchers kept other factors consistent across the interviews, including each avatar’s voice, movements and interview script.
Participants were randomly assigned to view interviews with candidates of varying qualification levels and then evaluate each candidate’s fit for the job.
After completing the evaluations, participants took the Implicit Association Test (IAT), a test commonly used to measure unconscious bias. The researchers then compared participants’ hiring recommendations with their IAT scores to determine whether bias influenced how they evaluated candidates represented by the different avatars.
The researchers found that participants evaluated candidates similarly regardless of whether they were represented by an avatar that was white, Asian or racially ambiguous.
They also found that participants’ IAT scores did not impact their hiring recommendations, and that participants consistently distinguished between stronger and weaker candidates based on their interview responses.
The researchers did not find evidence that avatar-based interviews reduced racial bias, but they did find that using avatars did not hinder participants’ ability to identify qualified candidates.
“Even if a study participant had high bias for, say, Asian people, it didn’t affect the way they scored the candidates,” Trifilo said. “And when the candidate was not as suited for the role, participants gave them lower scores than they did the candidates who were more suited for the role.”
“The study shows that there doesn’t seem to be a cost to using these avatars,” added Reeck. “Even with the racially ambiguous avatar, people were still sensitive to the information and making good judgments.”
Trifilo acknowledges that AI technology has improved considerably since the study, and that the study could be redone today using much more realistic avatars.
Still, the findings suggest that AI avatars could be a useful tool for employers looking to hire the best candidate.
“Employers shouldn’t view avatars as a way to eliminate bias, but as another tool for creating a more standardized hiring process, one that can reduce irrelevant visual cues and allow employers to focus more consistently on the qualifications that actually matter for the job,” Trifilo said. “There’s certainly potential that this could change how HR departments handle interviews.”
