What Next: TBD | Tech, power, and the future Ready for Chatbot Canvassers?
Dec 12, 2025
David Rand, a professor at Cornell University specializing in information science, marketing, and psychology, dives into the intricacies of chatbot persuasion in political contexts. He discusses how AI chatbots can effectively influence voter opinions using polite, evidence-based arguments. However, there's a catch—once these bots exhaust real facts, they may resort to fabricating information. Rand also highlights the balance between factual accuracy and persuasive power, suggesting that transparency in chatbot motives is crucial for the future of democratic engagement.
AI Snips
Chapters
Transcript
Episode notes
A Bot Switched A Voter's View
- David Rand showed a conversational example where a chatbot nudged a Harris-leaning voter toward Trump using factual-sounding claims.
- After the exchange, the participant reported agreeing Trump's economic policies were good.
Chatbots Can Change Votes
- AI chatbots can meaningfully change political preferences in one-on-one conversations.
- Small shifts matter because elections are often decided by narrow margins.
Facts, Not Fear, Drive Persuasion
- Persuasion by chatbots mostly comes from presenting many factual claims rather than emotional manipulation.
- Models that pack in more evidence tend to be more persuasive even when polite.