Me, Myself, and My AI: How artificial intelligence classification failures threaten consumers’ self-expression

Abstract

Drawing from AI classification experience and identity-based motivation frameworks, this research explores the impact of AI classification failures on consumers’ self-identification and examines how self-expression plays a crucial role in shaping these effects. Across five studies, this research reveals that AI classification failures can diminish consumers’ self-identification, increasing negative outcomes such as regret, defensive neurophysiological cues, and adverse downstream behaviors for service providers. Consistent with our identity-based account, the findings highlight that the detrimental impact of AI classification failures is more pronounced when consumers are motivated to utilize AI-recommended content for self-expression, and such effects are mitigated when these motives are not salient. This study offers important theoretical and managerial implications for the burgeoning field of service failures by shedding light on the nuanced impact of AI classification failures and their potentially detrimental effects. Finally, it provides managerial recommendations for recovering failures in AI classification experience.