The Dark Side of AI Anthropomorphism : A Case of Misplaced Trustworthiness in Service Provisions

Abstract

Anthropomorphism, the attribution of human-like traits, qualities, and mental states to non-human agents, is increasingly ubiquitous in artificial intelligence (AI) applications within service provisions. While imbuing anthropomorphism into AI agents enhances user interaction, it also has a dark side that poses significant ethical concerns by potentially causing harm to consumers. In response, our study constructs a framework to explain how anthropomorphic features in AI agents can lead to misplaced trustworthiness among consumers. We adopt a sociotechnical perspective and employ a case study design to achieve this. Our findings hope to advance the social sustainability of AI to promote responsible production and consumption patterns in global service markets.