Empowered yet unethical? Understanding GenAI misuse intention in generative AI chatbot use

Abstract

Purpose This study examines whether generative AI chatbot features are associated with user autonomy and competence and whether these psychological states are, in turn, associated with GenAI misuse intention, conceptualized here as an empowerment backfire mechanism. Design/methodology/approach Drawing primarily on the Self-Determination Theory, this study develops an empowerment-based model of GenAI chatbot misuse and examines the Neutralization Theory as an alternative account of moral rationalization. Data from 736 GenAI users were analyzed using structural equation modeling (SEM). Findings Results show that GenAI capability and accessibility were positively associated with autonomy and competence, which were positively associated with GenAI misuse intention. Situational anxiety moderated only the competence–GenAI misuse intention relationship. By contrast, neutralization mechanisms showed mixed associations, suggesting that empowerment-related perceptions may play a more central role than conventional rationalization in explaining ethically questionable GenAI use. Research limitations/implications Although the cross-sectional design limits causal inference, the findings highlight the unintended ethical consequences of empowerment-related psychological states in AI-enabled service settings. Originality/value This study introduces the empowerment backfire mechanism, showing that GenAI capability and accessibility may increase misuse intention through autonomy and competence.