Objectways uses Indian workers to train humanoid robots for three dollars hourly
Is this a scandal?
No longer — the story has resolved. Noise 5/100, cooling down, across 0 sources.
As the demand for physical AI and humanoid robotics accelerates, offshoring of physical data annotation to developing economies will grow rapidly, likely prompting calls for international labor standards for digital and physical data trainers.
Noise 5/100 — louder than 96% of tracked AI controversies.
Why it matters
The practice highlights the stark global economic disparities undergirding the AI supply chain, where low-wage workers train the very automation technologies that could eventually replace their jobs.
Key points
- Indian workers are being paid over three dollars per hour to wear head-mounted cameras and motion-tracking sensors while performing household chores.
- The physical data is collected by Objectways and subcontractor Qanat Consulting Services, which manages a pool of 2,000 workers.
- Workers operate in specialized environments like mock studio apartments and textile factories to record dozens of four-minute movement clips daily.
- Executives state the data will help humanoid robots learn human-like movements, aiming to automate repetitive physical tasks.
- The practice has reignited debates over labor exploitation and the long-term impact of AI automation on manual jobs in developing nations.
The story
Indian data annotators are earning approximately three dollars per hour to record daily household chores and manual tasks, providing critical training data for humanoid robots. Organized by the firm Objectways and its subcontractor Qanat Consulting Services, which manages roughly 2,000 workers, participants wear head cameras and motion tracking bands on their limbs while performing actions like washing dishes and cutting fruit. The captured motion data is designed to help humanoid robots learn human-like physical movements. While executives defend the practice as a stepping stone to free humans from mundane tasks, critics raise concerns over the low compensation and the potential for long-term job displacement, particularly as Morgan Stanley forecasts the global humanoid robot population to exceed one billion by 2050.
Who's involved
Argue that paying low wages to workers to train the systems that will eventually displace them is exploitative.
Collects human movement data from workers to train advanced AI systems for Fortune 500 clients.
Argues that mundane jobs must be automated so that humans can redirect their focus toward more useful, higher-value tasks.
Noise Level
The timeline
Labor practices in robot training revealed
Reports surface detailing how Indian annotators are paid roughly three dollars an hour to wear tracking gear and record chores for Objectways.
The forecast
As the demand for physical AI and humanoid robotics accelerates, offshoring of physical data annotation to developing economies will grow rapidly, likely prompting calls for international labor standards for digital and physical data trainers.
Forecast, not fact — an editorial estimate we score when this resolves.
That's the complete picture as of — nothing more to know right now. We'll update this page the moment it changes.
Join the Discussion
Discuss this story
Community comments coming in a future update
Be the first to share your perspective. Subscribe to comment.