Abstract:As the integration of intelligent systems within organizations becomes increasingly prevalent, there is a discernible shift from conventional human-to-human collaboration to a more cohesive form of collaboration that encompasses both human participants and intelligent systems. The existing body of literature has largely neglected to examine the impact of the characteristics inherent in human-intelligent system collaboration on employee innovative behavior. Utilizing the conservation of resources theory, this study proposes a moderated dual-mediation model to explore the dual mechanisms at play and their boundary conditions concerning innovative behavior. An analysis of data gathered from 414 employee questionnaires across 28 organizations reveals several key findings: (1) Employee work autonomy within the context of human-intelligent system collaboration significantly enhances employee innovative behavior; (2) Self-efficacy and hindrance stress function as mediators in the relationship between employee work autonomy in human-intelligent system collaboration and innovative behavior; (3) The relationship between human and intelligent systems positively moderates the effect of employee work autonomy in human-intelligent system collaboration on self-efficacy and hindrance stress, thereby amplifying the mediating roles of self-efficacy and hindrance stress in the relationship between employee work autonomy in human-intelligent system collaboration and employee innovative behavior. These findings elucidate the mechanisms through which employee autonomy in human-intelligent system collaboration influences innovative behavior, providing valuable insights for organizations seeking to implement intelligent systems to enhance human-intelligent system collaboration and support organizational change initiatives.