How Generative Artificial Intelligence Shapes Employees’ Responsible Innovation Behavior

The growing integration of generative artificial intelligence (GAI) into organizational innovation has raised critical ethical concerns, particularly regarding employees’ innovation-related conduct. Despite this trend, limited attention has been paid to how the inherent properties of GAI shape employees’ responsible innovation behavior (RIB). Addressing this gap, this study explores how key characteristics of GAI influence employees’ responsible innovation behavior (RIB). Using a mixed-method approach, we identify five ethically problematic GAI characteristics—hallucination, homogeneity, deficiency, unexplainability, and bias. Drawing on Social Cognitive Theory (SCT), we test a conceptual model with structural equation modeling (SEM). The results show that these characteristics weaken RIB indirectly by reducing GAI-driven moral efficacy (GME) and increasing GAI-driven moral disengagement (GMD). Furthermore, formal and informal punishment mechanisms moderate the effects of moral cognition on RIB. This study sheds light on the ethical risks of GAI-enabled innovation and offers managerial insights for promoting responsible use of generative technologies.