WHEN ARTIFICIAL INTELLIGENCE THREATENS CAREERS: A MODERATED MEDIATION MODEL OF JOB INSECURITY, EMOTIONAL EXHAUSTION, AND TURNOVER INTENTION

Abstract

The rapid integration of artificial intelligence into organizational operations has intensified employee concerns about job continuity, yet the psychological pathway linking such concerns to workforce withdrawal remains incompletely specified. Drawing on Conservation of Resources theory, this study examines how AI-induced job insecurity influences turnover intention, tests emotional exhaustion as a mediating mechanism, and evaluates perceived organizational support as a boundary condition attenuating the insecurity–exhaustion relationship. Using a quantitative, explanatory design, cross-sectional survey data were collected from 312 manufacturing employees selected through purposive sampling, and the moderated-mediation model was analyzed via partial least squares structural equation modeling (PLS-SEM) in SmartPLS. The measurement model satisfied reliability, convergent, and discriminant validity criteria. Results indicate that AI-induced job insecurity significantly increases turnover intention both directly and indirectly through emotional exhaustion, which functions as a partial mediator, confirming that perceived technological threat operates as a resource-depleting stressor driving defensive withdrawal. Perceived organizational support significantly weakens the insecurity–exhaustion relationship, substantiating a resource-substitution mechanism and helping reconcile prior inconsistent findings on the buffering role of support. The study extends Conservation of Resources theory by integrating its loss and gain dynamics within a single mechanism in the AI workplace and by treating AI-induced insecurity as a distinct construct.