How and when organizational AI adoption improves employee innovation performance

Artificial intelligence (AI) is being progressively embraced by service organizations. Nevertheless, its effect on employees’ innovative performance is still ambiguous. This study utilizes conservation of resources theory to construct a dual-path model that elucidates the contradictory effects of organizational AI adoption on employee innovation performance within the service industry. Using a mixed-methods approach, combining two field investigations (Studies 1 and 2) with a scenario-based experiment (Study 3) to examine our theoretical model. The results revealed that organizational AI adoption enhances innovation performance via perceived AI support while undermining it via perceived AI replacement. Moreover, AI trust functions as a pivotal boundary condition by strengthening the resource gain path and weakening the resource loss path. This study enhances the comprehension of human–AI collaboration by incorporating AI's augmentative and alternative features within a resource-based framework, providing actionable knowledge for service industries aiming to promote employee innovation performance through the adoption of organizational AI.