Robot training data market expands as startups pay workers to capture physical tasks

Traders Union
All News
Business & Wealth
Billionaires
Robot training data market

Highlights

Robotics startups like Microagi's Shift, Mecka AI, and Human Archive pay up to $30 an hour for data from more than 35,000 workers across over 30 countries, targeting $10 million in payouts by 2026.
Investor funding for robotics reached $33.4 billion globally in the first half of this year, already surpassing the $28.7 billion raised in all of 2025, as data demand rises tenfold.
Mordor Intelligence forecasts the AI training dataset market to grow from $11.9 billion in 2024 to $49.8 billion by 2031, but persistent margin pressure and uncertain data requirements challenge suppliers.

Global push to collect human task data

As first reported by Business Insider, companies are paying workers to record everyday physical activities, from cooking and laundry to warehouse operations and engine repair, to build datasets for robotics systems. The article describes how startups including Microagi's Shift, Mecka AI, and Human Archive are expanding data collection across homes, workplaces, and industrial settings as robotics developers seek far larger volumes of annotated real-world material.

Shift says it pays more than 35,000 people to record tasks in more than 15 countries, with participants able to earn up to $30 an hour and total payouts of around $10 million in 2026. Mecka AI says it has moved beyond ordinary video by developing headgear and sensor-equipped gloves that capture motion and force, while Human Archive says it now operates in more than 30 countries and pays between $1 and $10 an hour, or more, depending on the assignment.

Executives and researchers say the central constraint in robotics remains a shortage of training data. Encord says demand for robot training data is roughly 10 times higher than a year ago, with customer requests increasing from about 1,000 hours of annotated data to 10,000 or even 100,000 hours.

Funding surge brings labor and margin questions

Investor appetite for robotics is rising sharply as companies bet that better datasets, cheaper hardware, stronger AI models, labor shortages, and competition with China will speed commercial adoption. PitchBook data shows robotics companies raise $33.4 billion globally in the first half of this year, above the $28.7 billion raised in all of 2025.

That demand is also creating a new cross-border labor market with uneven economics. Workers in lower-cost countries can earn additional income by wearing recording equipment or remotely controlling robots, but critics say the competition for data tasks can suppress pay and leave workers with limited bargaining power, while some jobs are physically taxing and companies do not always explain in detail how consent is managed through local partners.

Longer term, the business case remains uncertain even if demand keeps growing. Mordor Intelligence projects the broader AI training dataset market will expand from $11.9 billion this year to nearly $49.8 billion by 2031, but industry participants say no one knows how much data robots will ultimately need, and simulation tools could reduce reliance on physical-world collection while thin margins already challenge some suppliers.

In our earlier coverage of DexForce’s approach to training humanoid robots, we explained how the company uses synthetic data generated in virtual simulations to reduce dependence on costly, time-intensive real-world human demonstrations. We noted that DexForce argues simulation can accelerate learning for a wide range of tasks and help address one of robotics’ main constraints: access to large volumes of high-quality training data.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.