When the Machine Judges: Employee Sensemaking of Procedural Justice in AI-Based Performance Evaluation Over Time

This study examines how employees make sense of procedural justice in AI-based performance evaluation over time. Although such systems are often framed as objective, employees frequently experience them as opaque, creating uncertainty about fairness. Drawing on sensemaking theory and procedural justice, we argue that fairness perceptions develop through repeated evaluation cycles rather than isolated events. Initial opacity may be tolerated as a technical limitation, but when it persists without explanation, employees may begin to question the fairness of the process. Using a qualitative, interpretive approach, we conduct semi-structured interviews with employees and HR managers to capture how these interpretations evolve over time. Narrative analysis is used to trace shifts in fairness perceptions across cycles and to identify distinct sensemaking pathways, including resignation, active questioning, and stabilization. We also examine how system changes, peer comparison, and organizational communication influence these trajectories