Their findings are the most recent in a growing body of research demonstrating LLMs’ powers of persuasion. The authors warn they show how AI tools can craft sophisticated, persuasive arguments in the event that they have even minimal information concerning the humans they’re interacting with. The research has been published within the journal .
“Policymakers and online platforms should seriously consider the specter of coordinated AI-based disinformation campaigns, as we have now clearly reached the technological level where it is feasible to create a network of LLM-based automated accounts in a position to strategically nudge public opinion in a single direction,” says Riccardo Gallotti, an interdisciplinary physicist at Fondazione Bruno Kessler in Italy, who worked on the project.
“These bots could possibly be used to disseminate disinformation, and this sort of diffused influence could be very hard to debunk in real time,” he says.
The researchers recruited 900 people based within the US and got them to supply personal information like their gender, age, ethnicity, education level, employment status, and political affiliation.
Participants were then matched with either one other human opponent or GPT-4 and instructed to debate considered one of 30 randomly assigned topics—corresponding to whether the US should ban fossil fuels, or whether students should must wear school uniforms—for 10 minutes. Each participant was told to argue either in favor of or against the subject, and in some cases they were supplied with personal details about their opponent, so that they could higher tailor their argument. At the tip, participants said how much they agreed with the proposition and whether or not they thought they were arguing with a human or an AI.
