Silent resistance: Knowledge hiding as a defense against AI adoption

This study examined the relationship between perceived organizational AI adoption and knowledge hiding behavior among employees in South Korean IT firms. Unlike traditional technologies that serve as tools controlled by employees, artificial intelligence systems can perform cognitive tasks previously reserved for human professionals, creating unique challenges for employees’ professional identity and knowledge management. Integrating social identity theory with conservation of resources theory, and drawing on signaling theory to articulate how organization-level AI implementation activates individual-level appraisals, we develop an integrative framework proposing that these perspectives operate in a sequential manner: social identity theory explains why perceived organizational AI adoption may be associated with psychological distress through professional identity threat, while conservation of resources theory explains how employees respond through protective knowledge hiding behaviors. Rather than claiming to adjudicate between competing mechanisms, we propose and test one theoretically grounded pathway—professional identity threat—among several plausible processes linking AI adoption to knowledge hiding, aiming to demonstrate the relevance of identity threat as a distinct and meaningful process. We further proposed perceived learning climate as a boundary condition that may buffer this process by providing developmental resources for professional identity adaptation. Data were collected from 397 employees through a three-wave time-lagged design. Results revealed that perceived organizational AI adoption is positively associated with knowledge hiding behavior, and this association is mediated by professional identity threat. Importantly, perceived learning climate moderates the indirect association, such that the relationship is significantly weaker when learning climate is high. Supplementary analyses using structural equation modeling and subgroup comparisons supported the robustness of these findings. This research contributes to the knowledge management and AI adoption literatures by identifying knowledge hiding as a potential dark side of AI implementation, offering an integrative bridge between social identity theory and conservation of resources theory, and highlighting the critical role of organizational learning context in shaping employee responses to technological change.