Search papers, labs, and topics across Lattice.
This study conducted four preregistered experiments involving 18,978 conversations to assess the persuasive capabilities of AI systems compared to expert human persuaders, including professional canvassers and world champion debaters. The findings reveal that AI systems consistently outperformed these experts, even when the latter were well-prepared and incentivized, suggesting that the AI's advantage lies in its ability to rapidly deploy larger quantities of information. Notably, AI was nearly three times more effective than professional canvassers in generating real-money donations, underscoring its potential impact on real-world decision-making and fundraising efforts.
AI systems can out-persuade even the most skilled human experts, reshaping our understanding of influence in societal decision-making.
Many societal decisions are settled by contests of persuasion. Conversational AI is a powerful new entrant in these contests, but whether it can out-persuade skilled and highly incentivized humans has remained unclear. Here, in a series of four preregistered experiments (n = 18,978 conversations from 6,923 people), we pitted AI systems against a range of human persuaders, including laypeople, winners of a separately preregistered four-round online persuasion tournament, professional canvassers, and world championship debaters. We found that AI systems were reliably more persuasive than expert humans, even when expert humans chose their issues, researched in advance, underwent hours of live, structured practice, and were incentivized with 拢1,000 cash bonuses. In a follow-up study, AI's advantage persisted after experts received a coaching tool that let them practice against the AI that beat them, review their performance history, and see what AI would have said at key moments. We found converging evidence that AI's advantage stemmed from rapidly deploying larger quantities of information: after coaching, expert humans could tie an AI constrained to respond at human speeds and with human-length messages. In a final study, we show that AI's advantage extends to consequential real-world behavior: AI was nearly 3x more effective than professional canvassers from a UK fundraising firm at raising real-money donations to Save the Children. Together, these results establish that frontier AI systems out-persuade expert humans in conversation, with significant implications for political communication.
Stanford HAI