The Cost of AI Adaptation: Territorial Control Loss Driving AI Loafing and Proactive Career Behavior

As artificial intelligence (AI) becomes deeply embedded in employees’ day-to-day work, many employees are expected not only to interact with AI intensively but also to work alongside AI systems that can learn from experience and continuously improve. Yet, the psychological costs of this “successful” AI integration remain underexamined. Drawing on conservation of resources theory and territoriality theory, we propose that intense AI interaction can ironically backfire when AI is perceived as highly adaptable, because employees may feel that a self-improving AI is increasingly encroaching on their discretion and ownership over valued work domains—what we term AI-caused loss of territorial control. We further theorize that this perceived loss of control prompts compensatory and defensive responses, manifested as greater AI loafing and stronger proactive career behavior. Moreover, we posit that career encouragement buffers these effects by mitigating the downstream consequences of territorial control loss. Based on data from a multi-wave field study and an experiment, we find support for our theorized model. Overall, our findings highlight a hidden cost of working with self-improving AI and identify career encouragement as a critical boundary condition for sustaining employee agency and adaptive career responses in AI-intensive work.