The AI training data market is exploding, and Micro1 is riding the wave. The four-year-old startup has expanded its gross annual run rate from $100 million to $500 million over the past eight months, according to a person familiar with the company.
Unlike AI model makers that compete for compute resources, a parallel boom is happening in the data layer. Companies like Micro1, Mercor, and Handshake hire domain experts—doctors, lawyers, scientists, and engineers—on a contract basis to label and annotate training data. Micro1 retains roughly 60% to 70% of its gross revenue, putting its net annual run rate between $150 million and $200 million.
The growth trajectory mirrors the broader AI infrastructure spending spree. While companies like Microsoft, Amazon, and Google pour billions into compute, researchers are beginning to hypothesize that future AI spending on data could rival or even exceed spending on compute. Some estimates suggest data costs could become the dominant factor in AI development budgets.
Micro1’s rapid expansion also highlights an emerging controversy in the AI data market. The company generates both human-annotated data and synthetic data—including automated video descriptions—that can be sold to multiple customers. This “off-the-shelf” data approach drives gross margins as high as 80% to 90%. However, critics argue that distributing such datasets to Chinese AI developers helps accelerate their models toward parity with top U.S. systems.
“We believe it’s shameful to claim American AI dominance desires while selling millions worth of data to countries that we are in adversarial competition with,” said Micro1 founder Ali Ansari last month on X, noting that unlike some competitors, Micro1 doesn’t sell data to Chinese model makers.
The startup originally began as an AI recruiting platform before pivoting to data labeling after noticing clients used its platform to recruit engineers for annotation work. Micro1 raised its Series A at a $500 million valuation last September and is reportedly in the process of raising another round at a significantly higher valuation.