How Romance Scams Became AI-Powered Fraud Operations
Online Safety·October 27, 2026
Romance scams used to be easy to spot: a clumsy message from a stranger, a sob story, and a sudden request for money. In an episode of Scientific American's Science Quickly podcast, host Rachel Feltman talks with criminologist Fangzhou Wang about how that picture has changed. Today's operations are more organized and more convincing, and they increasingly lean on artificial intelligence.
The conversation covers the tools scammers use to build trust. Dating apps and social platforms serve as the starting point, where fake profiles are designed to look credible. Deepfakes and AI-generated messages can keep a fabricated persona believable over weeks of conversation, and that slow buildup of intimacy is the heart of the con. When the request for money finally comes, it often takes the form of a cryptocurrency transfer, a payment method that is difficult to trace and hard to reverse.
Wang's research approaches these schemes as a sequence of steps rather than a single lie. Her work on pao fen operations, a term for the large-scale fraud networks behind many of these cases, uses what criminologists call a crime script, a stage-by-stage map of how the crime unfolds. Breaking the process into stages helps researchers identify the points where a potential victim could be warned, and it shows which prevention efforts are most likely to work. Her studies also look at the problem from the victim's side, since advice that ignores how people actually experience the deception tends to fall flat.
The broader lesson is that recognizing manipulation is a skill that can be taught before someone is already emotionally invested. Common warning signs include a new online partner who avoids video calls or in-person meetings, pushes to move the conversation off the dating platform, or steers talk toward investments and money. Anyone who has been targeted can also report the scam to local police and to the platform where it began, which helps investigators and warns others.