Effects of Anthropomorphism and Belief in Positive Machine Heuristics on Disclosure Intention in AI-Powered Digital Twin Cities: A Privacy Calculus Perspective

Artificial Intelligence (AI) is increasingly used for data collection, integration, and predictive analysis in large-scale public projects like digital twin cities—virtual replicas of real-world cities. However, privacy concerns remain, particularly when personal data is involved. While research indicates people may be more willing to disclose private information to AI systems than to humans, the privacy-related psychological mechanisms in human-AI interaction remain unclear. This study examines how anthropomorphism and belief in positive machine heuristics interact to influence personal data disclosure intention in AI-powered digital twin cities. Results from an online survey in Singapore (N = 1,000) reveal that belief in positive machine heuristics is associated with greater perceived benefits, lower perceived privacy risks, and greater disclosure intention. Anthropomorphism, however, showed a dual mechanism, associated with both higher perceived privacy risks and greater disclosure intention. Additionally, anthropomorphism is positively associated with perceived benefits, but only when belief in positive machine heuristics is low.