Your smart home camera company just asked you to pretend to steal packages for $2 a pop. Between December 18, 2024, and February 25, 2025, Anker’s Eufy division ran exactly this campaign—literally telling camera owners to “create events by pretending to be a thief” to train their AI detection algorithms.
The mechanics were elegantly dystopian. Upload theft videos via Google Form, submit PayPal details, collect your two bucks. Eufy aimed for 20,000 clips each of package thefts and car door-pulling scenarios. More than 120 users publicly participated, contributing to what became hundreds of thousands of training videos.
Privacy Theater Meets Data Hunger
This wasn’t Eufy’s first privacy stumble. According to security researchers, in 2023 their supposedly end-to-end encrypted cameras were caught streaming unencrypted footage to the web. Now they’re collecting massive video datasets while dodging basic transparency questions about storage, retention, or deletion policies. When pressed for details on total participation or data handling, Eufy declined to provide details.
The Gamification Continues
After the paid campaign ended, Eufy seamlessly transitioned users into their “Video Donation Program.” The app now features an “Honor Wall” showcasing top contributors—some submitting over 200,000 clips for digital badges and small gifts. It’s like Duolingo, but for training surveillance AI. They’ve expanded beyond doorbell footage to baby monitors, building proprietary datasets across your entire home.
Welcome to Surveillance Labor
This represents something new in the gig economy: turning homeowners into data laborers for surveillance capitalism. You get pocket change while companies build billion-dollar AI models from your daily life. The staged theft angle makes it feel like content creation, but you’re actually training systems that will monitor millions of homes.
The real question isn’t whether better theft detection is worth $2—it’s whether you understand what surveillance companies are building with your footage, and why they’re so reluctant to explain their data practices when the cameras stop rolling.
