Rewritten Article
To score a complimentary full house cleaning at no cost, customers agree to let startup Shift collect video data captured via a camera-equipped headset worn by the human cleaner completing the job. This raw footage is then licensed to third-party teams working to build and train AI-powered household robots.
Shift is owned by MicroAGI, a German data research laboratory founded in 2025. In its early stages, the company hired contract workers to wear its specialized camera-enabled caps and record themselves completing routine household tasks. This style of on-the-ground data collection has already been active for months across the United States, Germany, Turkey, and a number of other European countries. Notably, Shift itself has no plans to produce the AI robots that its dataset will inform.
Most recently, Shift rolled out its public cleaning service, offering free home cleans as a tactic to get its data-collecting workers into ordinary people’s private residences. The service is currently available in New York City and across most of Europe. The company has marketed its cleaning role as an accessible side hustle for college students, advertising it as the “best work from home side gig.”
“We are completely transparent about our process,” said Anton Poletaev, co-founder of MicroAGI and co-CEO of Shift. “We don’t hide that we are collecting your data — the difference is that you actually get rewarded for sharing it here, and no one is lying to you about what’s happening.”
But even with this explicit honesty, many customers may still not fully grasp the full implications of handing over this level of intimate personal data. Shift’s business model also raises pressing questions about what this AI-driven future means for the human workers who are actively involved in training their own eventual replacements.
Rewritten Article
