Exploring users' post-adoption use of generative AI: An attitudinal ambivalence perspective

Abstract

As generative AI (genAI) has advanced, the intricate interplay of its technical potential and ethical perils has become more pronounced, fostering a growing ambivalence in users' attitudes towards genAI technology. Drawing upon the attitudinal ambivalence perspective (i.e., the simultaneous occurrence of positive and negative evaluations of genAI use) and cognitive appraisal theory of emotion, our study proposes and tests an integrative research model to understand how users' attitudinal ambivalence towards genAI technology navigates their negative and positive emotional responses and shapes their post-adoption behaviors. We surveyed 530 genAI users and employed the structural equation modeling approach to test our research model. We find that attitudinal ambivalence is significantly associated with users' extended use and avoidance through the mediation of user trust and fear. Additionally, transparency significantly moderates the effects of attitudinal ambivalence on user trust and fear. Our study advances nature and consequences of attitudinal ambivalence towards genAI and provides insights for practitioners contemplating deploying genAI.