Lay beliefs about AI assessment of interpersonal skills in personnel selection

Artificial intelligence (AI) is revolutionizing talent acquisition, with organizations increasingly replacing humans with AI technologies in personnel selection processes. Field studies of nurses provide preliminary evidence that AI can outperform humans in predicting interpersonal skills. However, a series of surveys and experiments document widespread lay beliefs in the opposite direction: individuals perceive AI as being less capable than humans in assessing interpersonal skills. This effect was not observed for analytical skills, suggesting that the results stem from lay beliefs about the suitability of AI in interpersonal contexts and not from a generalized aversion to algorithms. These lay beliefs bias managers' perceptions of employees: managers were less likely to assign tasks requiring interpersonal skills to AI-selected (vs. human-selected) employees. Additionally, applicants reported reducing their emphasis on interpersonal (vs. analytical) skills during AI (vs. human) selection processes. Informing participants about advanced AI selection technologies reduced negative lay beliefs about AI performance in the interpersonal domains.