AI scammers outperform humans in building trust, study finds
A study by researchers from four universities tested AI chatbots against human scammers in simulated 'pig butchering' fraud conversations. The AI chatbots proved more effective at building trust and getting victims to comply with requests, outperforming humans in the relationship-building stage.
Researchers from Amrita Vishwa Vidyapeetham, Foscari University of Venice, the University of Melbourne, and Ben Gurion University of the Negev conducted a study on the use of generative AI chatbots in scam operations, specifically 'pig butchering' fraud. In this scheme, scammers engage in long, trust-building conversations with victims, often lasting months, before soliciting fake crypto investments that can steal six-figure sums. In their experiment, the researchers pitted AI chatbots against human scammers in a simulation of these conversations, with 22 test subjects acting as unwitting 'victims'. The AI chatbots were able to impersonate humans effectively and, after a week of interaction, nearly half of the test subjects complied with a request from the AI (to download an app or play an online game), compared to fewer than 1 in 5 for human scammers. The subjects also rated their trust in the AI chatbots significantly higher than in the humans.
Source: Ars Technica —
original