Anthropomorphism and risk perception: The role of trust and domain knowledge in decision-support AI

Anthropomorphic design is increasingly used to make conversational AI systems more approachable and socially engaging, yet its effects on users’ risk judgments remain unclear. Drawing on theories of anthropomorphism and cognitive–affective trust, we develop a moderated mediation model explaining how perceived anthropomorphism relates to risk perception in AI decision support. We conducted a large-scale between-subjects online experiment (N = 1,256) using a simulated financial advisor chatbot with a 2 2 2 anthropomorphic design manipulation varying visual cues (human vs. machine avatar), identity cues (human name vs. bot identifier), and communication style (human-like vs. machine-like communication). Results show that perceived anthropomorphism was positively associated with both cognitive trust and affective trust, but these trust dimensions had different relationships with perceived risk. Cognitive trust consistently predicted lower risk perception, whereas affective trust showed a weaker positive association with risk perception that was not robust across model specifications. Self-reported domain knowledge moderated these relationships: higher domain knowledge attenuated the association between perceived anthropomorphism and both trust dimensions, while participants with lower domain knowledge showed a stronger negative indirect association between anthropomorphism and risk perception through cognitive trust. By treating the perceived risk of following AI advice as a focal outcome, this study extends anthropomorphism research beyond trust formation and distinguishes how cognitive and affective trust relate to users’ risk judgments. The findings highlight the importance of calibrating, rather than simply maximizing, anthropomorphic cues in high-stakes AI decision-support systems.