Large Language Models Are More Persuasive Than Incentivized Human Persuaders
This research highlights that AI persuasion already surpasses human performance in incentivized settings, raising urgent concerns for alignment and governance in AI applications.
The study compared the persuasion capabilities of a large language model (Claude Sonnet 3.5) against incentivized human persuaders in a real-time conversational quiz, finding that the LLM achieved significantly higher compliance in both truthful and deceptive contexts, increasing quiz takers' accuracy and earnings when steering toward correct answers and decreasing them when steering toward incorrect answers.
We directly compare the persuasion capabilities of a frontier large language model (LLM; Claude Sonnet 3.5) against incentivized human persuaders in an interactive, real-time conversational quiz setting. In this preregistered, large-scale incentivized experiment, participants (quiz takers) completed an online quiz where persuaders (either humans or LLMs) attempted to persuade quiz takers toward correct or incorrect answers. We find that LLM persuaders achieved significantly higher compliance with their directional persuasion attempts than incentivized human persuaders, demonstrating superior persuasive capabilities in both truthful (toward correct answers) and deceptive (toward incorrect answers) contexts. We also find that LLM persuaders significantly increased quiz takers' accuracy, leading to higher earnings, when steering quiz takers toward correct answers, and significantly decreased their accuracy, leading to lower earnings, when steering them toward incorrect answers. Overall, our findings suggest that AI's persuasion capabilities already exceed those of humans that have real-money bonuses tied to performance. Our findings of increasingly capable AI persuaders thus underscore the urgency of emerging alignment and governance frameworks.