Shame or Anger? The Impact of Negative Performance Feedback Sources (AI Versus Leader) on Employees' Job Crafting

With the growing adoption of artificial intelligence (AI) in organizational performance management, AI feedback has increasingly supplemented or replaced leader-delivered evaluations. While prior research has addressed issues of fairness and accuracy in AI assessments, relatively little is known about how employees emotionally and behaviorally respond to negative performance feedback (NPF) from different sources. Building on Affective Events Theory, this study investigates how leader versus AI elicits distinct emotions, shame and anger, and how these emotions subsequently influence employees' job crafting. Two studies were conducted to test the proposed model. Study 1 employed a scenario-based experiment to compare employees' emotional reactions. Results indicate that leader NPF evokes greater shame, whereas AI NPF induces stronger anger. Study 2 used survey data from nine enterprises in China to further test the underlying mechanisms. Results show that shame and anger mediate the effects of leader and AI NPF on promotion-oriented and prevention-oriented job crafting, respectively. Moreover, leader trust weakens the relationship between leader NPF and shame, while algorithm aversion strengthens the relationship between AI negative feedback and anger. This study advances understanding of the emotional mechanisms underlying employees' responses to negative feedback and offers practical insights for designing effective human-AI feedback systems in organizations.