AI Interviews Push Candidates to Exaggerate Qualifications
A new study reveals that job seekers often exaggerate their skills when interviewed by AI, as the technology fails to distinguish between authentic and embellished responses.
Job candidates are increasingly finding themselves face-to-face with artificial intelligence rather than human recruiters during initial screening. While this shift promises efficiency for companies, a recent study from the University of Georgia suggests it creates a strange dynamic where applicants feel compelled to overstate their abilities. The research indicates that when people know a machine will judge them, they are more likely to embellish the truth, a behavior that human interviewers would typically penalize.
The study, published in Information Systems Research, highlights a significant gap in how AI evaluates candidates compared to humans. Unlike human reviewers who can often detect inauthenticity, the AI systems tested did not penalize applicants who stretched the truth. This discrepancy means that candidates who exaggerate their qualifications may score just as highly as those who are honest, creating a competitive disadvantage for genuine applicants.
AI Systems Miss Subtle Deception
According to Akshat Lakhiwal, an assistant professor at the University of Georgia’s Terry College of Business, the core issue is that current AI rating systems cannot tell when a candidate is stretching the truth. In a one-way video interview, applicants record answers to standard questions. When told their responses would be graded by an algorithm, participants reported a considerable increase in deceptive embellishments. They described feeling helpless and disoriented, leading them to throw the kitchen sink at the situation to please an unpredictable evaluator.
Human evaluators, by contrast, seemed able to discern these behaviors. In side-by-side comparisons, humans generally gave lower ratings to candidates who appeared inauthentic and higher ratings to those who engaged in more genuine behavior. This difference suggests that the AI’s inability to detect nuance or intent may inadvertently reward exaggeration, altering the playing field for job seekers who are trying to be honest.
Transparency Reduces the Need to Exaggerate
The researchers found that explaining the evaluation process to candidates significantly curtailed their tendency to exaggerate. In an experiment, one group was told their videos would be reviewed by AI and then given specific details about what the system looked for, such as facial expressions, verbal sentiment, and specific keywords. This group reported and displayed the same amount of authentic behavior as those who were told they were being reviewed by a human. Telling applicants more about the process allows them to be more authentic, Lakhiwal noted.
The catch, however, is that many companies do not provide this level of transparency. Candidates often do not know how the AI works, leading to anxiety and a desire to game the system. As reported by GN technics/ai (en-US), this lack of clarity forces applicants to guess what the machine wants, often resulting in overcompensation. The study suggests that for AI interviews to be fair, companies must be more open about their criteria.
Candidates Feel Helpless Without Clarity
Many job seekers feel they have no choice but to participate in AI-driven interviews, even when applying to their dream companies. The uncertainty about how the system evaluates them creates a stressful environment where authenticity is difficult to maintain. When people do not understand the criteria, they tend to act in ways they imagine the machine prefers, rather than presenting their true selves. This behavioral shift affects the outcome of the interview, potentially skewing hiring decisions toward those who are better at performing for an algorithm.
The trade-off for companies is that while AI saves time and money, it may compromise the quality of the screening process. If the system cannot distinguish between a qualified candidate and one who is merely good at exaggerating, the hiring manager may end up with a pool of applicants who do not accurately reflect their true capabilities. The study underscores the need for a more transparent and human-centric approach to AI-assisted hiring to ensure fairness and accuracy.
