AI Job Interviews Push Candidates to Exaggerate, Study Finds

AI Job Interviews Push Candidates to Exaggerate, Study Finds

Job seekers facing automated video interviews are significantly more likely to embellish their qualifications than when speaking with human recruiters, according to new research from the University of Georgia's Terry College of Business.

The study, which examined hundreds of online job seekers, found that candidates reported and displayed considerably more "deceptive embellishments" when they knew artificial intelligence would evaluate their responses. More troubling: the AI systems currently used by employers failed to detect this behavior, rating exaggerating candidates just as highly as truthful ones with equivalent qualifications.

Why it matters

As companies accelerate adoption of AI screening tools to reduce hiring costs, they may be inadvertently selecting for candidates willing to stretch the truth rather than those with genuine qualifications. The findings suggest current automated systems lack the discernment human evaluators bring to interviews, creating a blind spot that could undermine hiring quality.

The transparency solution

The research team, led by assistant professor Akshat Lakhiwal, identified a straightforward remedy: explaining how the AI evaluation works.

When candidates received specific details about the screening process—that the system would analyze facial expressions, verbal sentiment, keywords, and rate responses based on teamwork, abilities, work style and personality—their behavior changed. This informed group displayed the same level of authentic behavior as candidates told they would be reviewed by humans.

"Traditionally, companies have refrained from transparency in the hiring process," Lakhiwal noted. "They don't like that word because they feel if participants know how they will be evaluated, the applicants may game the system. But here we found that telling applicants more about the process allows them to be more authentic."

How candidates respond to AI uncertainty

The shift in behavior stems from candidates feeling disoriented by opaque automated systems, according to the research. When applying to desired positions, job seekers feel they have little choice but to participate in AI interviews, yet they don't understand how these systems work.

"They seemed to be throwing the kitchen sink at the situation to try to give the 'evaluator' what it was looking for," Lakhiwal said.

When asked whether embellishing qualifications was ethical, participants said they considered it necessary to perform well in an unpredictable environment.

Human evaluators who reviewed the same video interviews generally penalized candidates who appeared to exaggerate, giving higher ratings to those displaying more authentic engagement—a nuance the AI systems missed entirely.

Implications for hiring technology

The findings highlight a critical gap as organizations deploy AI to streamline recruitment. While automated screening can reduce scheduling friction and enable faster comparisons across candidates, current systems lack the ability to detect behavioral cues that human recruiters use to assess authenticity.

The study tested an industry-favored AI evaluation system against human judgment, revealing that what companies gain in efficiency they may lose in assessment accuracy.

Lakhiwal emphasized that transparency doesn't require revealing proprietary algorithms. Candidates need to understand the evaluation process conceptually—the same way they understand what being interviewed by a human means—not the technical details of which models analyze their responses.

The study was published in Information Systems Research. Co-authors include Che-Wei Liu of Arizona State University, Hillol Bala of Indiana University, and Hung-Yue Suen of National Taiwan Normal University. The research was first reported by the University of Georgia.