Based on social learning theory and self-determination theory, this study investigates the dual-pathway influence mechanism of leaders' AI symbolization on employee task performance. It also examines the moderating role of leader–member exchange (LMX) in shaping these effects.
We conducted two studies: a scenario-based experiment (N = 243) and a multi-wave field survey (N = 379) with matched leader–employee pairs. Regression analyses, structural equation modeling, and bootstrapping were employed to test our hypotheses.
Leaders' AI symbolization has a double-edged effect on employee task performance. It enhances performance by increasing perceived AI utility and encouraging promotion-oriented task crafting, but it also heightens job insecurity and prevention-oriented task crafting, which affect performance through a different pathway. High-quality LMX strengthens the positive pathway and buffers the negative pathway.
This study uniquely integrates dual pathways to clarify how leaders' AI symbolization impacts employees, highlighting LMX as a crucial relational boundary shaping employee responses to AI-driven workplace changes. Beyond its theoretical contributions, the study offers practical guidance for organizations by suggesting ways to foster employee adaptability and proactive behaviors in AI-enabled workplaces.
