How and When Leader Anthropomorphizing AI Shapes Employee–AI Collaboration: An Identity Perspective

As artificial intelligence (AI) increasingly enters organizations as coworkers, it reshapes identity processes by redefining who employees are and who AI is within the workplace. Yet, how leaders shape this identity reconstruction and its implications for employee-AI collaboration remains largely unknown. Drawing on identity theory, this study examines how leader anthropomorphizing behavior toward AI (LABA) influences employee-AI collaboration through two identity pathways. We propose that LABA simultaneously triggers employee identity threat by undermining role meaning and subjectivity, while also enhancing perceived AI social presence by positioning AI as an agentive social actor. These opposing processes lead to divergent collaboration outcomes. Moreover, we identify employees’ self-concept clarity (SCC) as a key boundary condition. Employees with lower SCC are more susceptible to identity threat in response to LABA, whereas those with higher SCC are more likely to perceive stronger AI social presence. Evidence from a three-wave survey (N=416) and an online scenario experiment (N=164) supports our model. Overall, this research highlights LABA as a double-edged leadership practice that reshapes both employee identity and AI identity, advancing understanding of identity dynamics in AI-enabled workplaces.