Scope and limits of AI fundraisers: Moderated serial multiple mediation model between artificial emotions and willingness to donate via humanness and empathy

Charitable organizations are experiencing a global labor shortage. Artificial intelligence (AI) chatbots have been employed to help combat the shortage, however, the scope and limits of AI fundraisers are still unclear, calling for further investigation. To improve our understanding of AI fundraisers, this study recruited 654 adults from an online crowdsourcing platform and developed six chatbot agents (emotional vs. factual × no image vs. machinelike image vs. human image). First, we employed an independent samples t-test to examine the effect of chatbots' conversational style on willingness to donate to Ukraine war victims. This study also tested the mediating roles (independent and serial) of perceived humanness and empathy toward victims in the relationship between chatbots' emotional expression and willingness to donate by conducting a serial multiple mediation analysis. Finally, this study tested the moderating role of visual cues (no image vs. machinelike image vs. human image) by conducting a moderated serial multiple mediation analysis. The results revealed that emotional chatbots yield a higher willingness to donate, and perceived humanness and empathy toward victims mediate this relationship independently and serially. However, the results suggest that visual cues did not significantly moderate the relationship between chatbot agents' emotional expression on willingness to donate.