aceModerate evidence
More
Charity chatbot's compassion raised giving intent; its "many others donate" pitch backfired
Study title: How Should AI Agents Ask for Donations? Comparison of Emotional and Normative Appeals in AI-Mediated Fundraising
- Evidence type
- Controlled or validated comparison
- Sample
- n = 206
- Sources
- 1 sourcePeer-reviewed
- Published
- 2 min read
- Truvace record
- TRV-2026-2581cite or check it
Study title: How Should AI Agents Ask for Donations? Comparison of Emotional and Normative Appeals in AI-Mediated Fundraising
A between-subjects experiment with 206 participants tested how a charity AI agent should ask for money. Each person interacted with an agent that used either an emotional appeal, with messages evoking compassion and empathy, or a normative appeal, stressing that many others donate.
Read the study abstract
AI chatbots are increasingly being used by charities to engage potential donors. However, as people may exhibit algorithm aversion and heightened scrutiny toward AI, it remains unclear whether persuasive strategies established in human fundraising operate similarly when delivered by an AI agent. In a between-subjects experiment (N=206), participants interacted with a charity AI agent that applied either an emotional appeal (messages evoking compassion and empathy) or a normative appeal (messages emphasizing that…
55Gain side
How the two sides scoreClose call
55Problem side
The gain
In an experiment with 206 participants, a charity AI agent's compassion-based messages raised donation intentions by inducing empathy.
The problem
In the same 206-person experiment, the AI agent's "many others donate" messages made the pitch less believable, weakened anticipatory guilt and lowered donation intentions.
A fixed formula sets both scores from the evidence and its sources. Reader votes don’t move them.
What the scores mean
G is the gain side, out of 100. It goes up when the gain is specific, when it measurably helped people and when the sources behind it are strong.
P is the problem side, out of 100. It goes up when the problem is specific, when it measurably hurt people and when the sources behind it are strong.
In brief
The emotional appeal significantly raised donation intentions by inducing empathy. The normative appeal worked against itself: it reduced how believable the messages were, which weakened anticipatory guilt and ended up lowering donation intentions.
Charities increasingly use AI chatbots to engage potential donors, and the researchers point to algorithm aversion and heightened scrutiny of AI as reasons that tactics proven in human fundraising may not carry over. The results make message believability a boundary condition for AI-delivered normative appeals.
Main points
- The study was a between-subjects experiment with 206 participants who interacted with a charity AI agent.
- One group received emotional appeals evoking compassion and empathy; the other received normative appeals saying many others donate.
- Emotional appeals significantly increased donation intentions, with empathy as the mechanism.
- Normative appeals lowered message believability, which weakened anticipatory guilt and reduced donation intentions.
- The authors identify believability as a critical boundary condition for AI-delivered normative appeals and call for design that calibrates message strategies.
The full story
The research starts from a gap: charities are adopting AI chatbots to engage donors, but it is unclear whether persuasion strategies from human fundraising work the same way when an AI delivers them, given that people may show algorithm aversion and extra scrutiny toward AI.
Participants were split between two conditions. In the emotional condition the agent used messages meant to evoke compassion and empathy. In the normative condition it emphasised that many others donate. Outcomes were donation intentions and the psychological steps leading to them.
Emotional appeals raised donation intentions through empathy. For normative appeals the chain ran the other way: lower believability, then weaker anticipatory guilt, then lower donation intentions. The authors conclude that AI agents soliciting donations need carefully calibrated message strategies.
What this doesn’t fix
The finding measures donation intentions, not actual donations, in a single experiment with 206 participants.
Sources
1 on file- International Journal of Human–Computer Interactiondoi:10.1080/10447318.2026.2719117
Talk it over
This record found a gain and a problem in the same place. Which matters more to you, and who might answer differently?
Charity chatbot's compassion raised giving intent; its "many others donate" pitch backfiredhttps://truvace.com/item/how-should-ai-agents-ask-for-donations-comparison-of-emotional-and-normative-app
This record found a gain and a problem in the same place. Which matters more to you, and who might answer differently?
The gain In an experiment with 206 participants, a charity AI agent's compassion-based messages raised donation intentions by inducing empathy.
The problem In the same 206-person experiment, the AI agent's "many others donate" messages made the pitch less believable, weakened anticipatory guilt and lowered donation intentions.
Answer in the threadHow do we know? What can this kind of evidence show, and what can’t it?
The evidence This rests on a controlled comparison with a sample of 206, from International Journal of Human–Computer Interaction.
Answer in the threadWhat does it leave open? What else would you want to know before acting on it?
The record’s own caveat The finding measures donation intentions, not actual donations, in a single experiment with 206 participants.
Answer in the threadHave you seen AI used this way where you study, work or live? Did it go the same way?
Answer in the thread
The questions are the same on every record. The details under them are taken from this record as written.
What readers say