The incident occurred while Gore-Kormanik investigated a platform called Dora, one of roughly 28 fraudulent applications identified as part of an industrial-scale catfishing operation. The scale of the deception recently came to light after Anthropic detected anomalous activity involving a prepaid account executing over 100,000 API requests daily via its Claude model.
Industrial-scale AI catfishing hits the dating app market
Security researcher Matthew Gore-Kormanik received a video call from a persona named Jennifer, complete with a bio and a voice that claimed to hear his own—despite his microphone being disconnected. This encounter exposed a sophisticated network of fraudulent dating apps fueled by autonomous AI agents designed to deceive users.
Anthropic threat intelligence researcher Chris Cronbaugh detailed the scheme at the Sleuthcon cybersecurity conference, explaining that these platforms operate with minimal human intervention. Instead, the apps deploy autonomous AI personas to maintain persistent, convincing conversations with victims. The findings, now documented in Anthropic's report on AI misuse, reveal a shift toward automated, high-volume social engineering where the illusion of human connection is generated entirely by machine learning.
Comments (0)
No comments yet. Be the first!