AI Adoption as Screening Design: When Candidate Choice Becomes Signal, by Jonathan H. Westover PhD

Abstract: This article examines how firms should integrate artificial intelligence into labor-market screening when applicants can choose between human and AI interviewers. Drawing on a natural field experiment involving 70,000 job applicants and recent theoretical advances in mechanism design, we show that AI adoption is fundamentally a design problem rather than a simple substitution decision. When applicants select their preferred interviewer, this choice itself becomes an informative signal about underlying abilities—a phenomenon we term "choice-as-signal." The welfare implications depend critically on whether firms incorporate this signal into hiring decisions and whether applicants anticipate such use. Evidence suggests that hybrid screening systems combining human and AI evaluation outperform either technology alone, and that specialized assignment—matching each screener to the dimensions they assess most accurately—can improve match quality. These findings challenge conventional automation narratives and reveal novel trade-offs between worker autonomy and information revelation in AI-augmented hiring.

See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.