Abstract
A human-centric understanding of purchase intention in generative AI contexts requires attention to how users interpret and experience AI systems under conditions of algorithmic opacity. Rather than treating AI adoption as the outcome of attribute evaluation, we argue that anthropomorphic cues shape how users construe the system itself. When AI is experienced as socially responsive rather than purely instrumental, evaluative emphasis shifts toward trust beliefs and interactional arousal. Drawing on Computers as Social Actors (CASA) and Mind Perception Theory, we theorize that anthropomorphism restructures the experiential architecture through which intention is formed. Using structural equation modeling with data from 526 generative AI users, results show that anthropomorphism influences purchase intention both directly and indirectly through trust and arousal, with relational mechanisms exerting stronger effects. By foregrounding user construal and experiential engagement, this research advances a human-centric framework for understanding AI adoption in opaque, interactive environments.
Recommended Citation
Cheng, Jao-Hong and Wang, Li-Ming, "Anthropomorphism and Relational Ontology of Purchase Intention in Generative AI Use: Insights from Mind Perception Theory and the Computers as Social Actors Paradigm" (2026). ICEB 2026 Proceedings (Beijing, China). 15.
https://aisel.aisnet.org/iceb2026/15