AI-Powered Virtual Influencers: Authenticity, Warmth, and Donation Intention in a Single-Stimulus Charity Appeal Context

This study examines how perceived authenticity and warmth are associated with parasocial interaction (PSI) and donation intention in response to an AI-powered virtual influencer charity appeal. Rather than claiming that virtual influencers are broadly more effective than human influencers, the study focuses on a bounded setting in which respondents evaluated one mimic-human virtual influencer post. Drawing on parasocial interaction theory and selected insights from self-determination theory, the model proposes that authenticity functions as a credibility cue, warmth functions as a benevolence cue, and PSI captures the resulting quasi-social bond. Using survey data from 240 respondents, the findings support significant links from perceived authenticity to warmth and PSI, and show that warmth and PSI are associated with stronger donation intention, including a significant sequential indirect effect. The contribution is therefore more modest but more precise: the study clarifies how audiences interpret a single virtual influencer charity appeal and identifies warmth as an important emotional mechanism in nonprofit communication. The findings offer a foundation for future experimental research comparing AI and human influencers, alternative virtual influencer types, and actual donation behavior.