Deepfake model farms exploit vulnerable individuals

A disturbing trend is emerging: individuals are being lured into becoming unwitting models for deepfake Technology, often promising quick cash. A recent Wired investigation reveals a disturbing ecosystem where people are being recruited to generate the very data fueling increasingly sophisticated scams.

The lure: high pay, demanding hours

The advertisements, primarily found on Telegram channels, offer enticing terms – up to 100 video calls per day, and sometimes escalating to 150. These aren't legitimate modeling gigs; they're data harvesting operations. Individuals are tasked with providing a constant stream of video and audio recordings, often for extended periods, in exchange for relatively low pay.

The process involves downloading software to manipulate facial features using AI. The goal? To create realistic digital replicas for use in romance scams, a particularly insidious form of fraud where perpetrators cultivate emotional connections with victims before extracting money or personal data.

The locations with the highest concentration of these opportunities include Turkey, Russia, Ukraine, and Belarus, alongside various Asian countries. While the Telegram channels rarely explicitly mention