hotai.news
← Back to AI Newsflash
AI News / Event details
chinatalk.media · Published

ChinaTalk investigates the phenomenon of US data companies selling reinforcement learning environments to Chinese AI labs

Automated summary · Based on collected summaries; not individually verified
Prepared by HotAI

What happened

ChinaTalk's investigation reveals that at least nine US data companies, including Mercor, Surge AI, Turing, and AfterQuery, have sold reinforcement learning environments and other data worth approximately $500 million to Chinese labs such as Alibaba, Tencent, and ByteDance. This assistance has helped improve capabilities in coding, office work, and other fields through post-training.

Background

What is reinforcement learning? Actions, feedback, and long-term rewards

Reinforcement learning adjusts a policy through interactions and rewards, typically to improve cumulative return. Rewards need not provide a correct answer at every step, and actions can change future states. Unlike learning only from fixed labeled examples, it depends on interaction. A poorly designed reward can produce high scores without the intended outcome.

Reading notes

The event summary comes from collected excerpts. Background passages quote existing HotAI articles to explain terms and do not establish new facts about this event. Check the original sources for product results, conditions and subsequent updates.

Original source

Sources and collection timeline

SourceReport titleOriginal publication timeCollected by HotAI
chinatalk.mediaChinaTalk investigates the phenomenon of US data companies selling reinforcement learning environments to Chinese AI labs2026-11-11 19:072026-11-11 20:47