A Moderated Mediation Model of AI-Driven Identity Threats and Employee Cyberloafing: The Role of AI-Inclusive Identity

Abstract

This study intended to examine how human–AI collaboration-based identity threat appraisals in the form of the loss of autonomy and loss of skill trigger a professional identity that fosters cyberloafing. Based on social identity theory, this study applied a three-wave survey design with 507 employees. The proposed research model was tested using partial least squares structural equation modeling (PLS-SEM) with SmartPLS 4, which enabled the assessment of both measurement and structural models. The perceived loss of skill and loss of autonomy are positively associated with professional identity threat, which mediates their relationships with cyberloafing. AI-inclusive identity weakens these associations for the loss of autonomy, suggesting that employees with strong AI-inclusive identity exhibit weaker professional identity threat. When integrating AI, organizations should mitigate appraisals of the loss of autonomy and loss of skill through participatory design, role redesign, and communication that emphasizes unique human contributions. Supporting healthy AI–human identity integration may reduce counterproductive behaviors such as cyberloafing. By positioning identity threat appraisals as human–AI collaboration-driven antecedents of professional identity threat and cyberloafing, this study extends social identity theory to human–AI contexts. It further demonstrates that over-identification with AI may heighten professional identity threats by diminishing the value of uniquely human contributions.