Moral Accountability in the Face of Human and AI Misconduct: Moral Emotions and Consumer Punishment of Companies
(A2026-133112)

AUTHORS

Jacqueline Veith, Johannes Gutenberg-Universit�t Mainz; Frank Huber, Johannes Gutenberg University Mainz

ABSTRACT

This research investigates how responsibility attribution and moral emotions influence consumers’ behavior following errors by human employees or AI agents. Across two scenario-based studies, we examined (1) investment fund errors by human or AI, and (2) subscription-cancellation errors by non-anthropomorphic versus anthropomorphic chatbots. Results indicate that agent type does not directly affect negative word-of-mouth (nWOM), instead, responsibility attributed to the agent and moral emotions mediate this relationship. Human errors elicit stronger responsibility and more intense moral emotions than AI errors, leading to greater nWOM. Anthropomorphized AI is attributed a higher mind, reducing corporate responsibility, while responsibility attribution to AI is already triggered by anthropomorphism. These findings highlight the importance of mediators in consumer reactions to AI errors, offer practical implications for handling them, and guide future research on AI and moral emotions.