Chinese Robotics Startup Accuses OpenAI of Model Distillation
Ant-backed JoyIn's CEO claims OpenAI lifted from his firm's work, flipping the usual distillation accusation script.
The AI intellectual-property wars just got a new twist. JoyIn, a Chinese humanoid robotics startup backed by Ant Group, is now pointing fingers at OpenAI — not the other way around. The company's CEO published an article in Chinese raising pointed questions about suspicious similarities between JoyIn's new robotics model and OpenAI's own publications.
This is a bold move. The term "distillation" — using a powerful model's outputs to train a cheaper, smaller one — has been weaponized mostly against Chinese AI firms. DeepSeek faced those exact allegations earlier this year. Now JoyIn is flipping the script, suggesting the knowledge transfer may have run in the opposite direction.
For traders watching the AI space, this drama matters. Every IP dispute chips away at the narrative that Western labs hold an unassailable moat. If a Chinese robotics startup can credibly claim OpenAI borrowed from its work, that changes the competitive calculus — and the valuation story for anyone long on AI infrastructure plays.
The claim hasn't been independently verified, and OpenAI has not publicly responded based on available reporting. But the fact that a relatively unknown startup is willing to go public with this accusation signals growing confidence among Chinese AI developers that they have something worth protecting — and the clout to say so.
This story is still developing fast. Continue reading at US Top News and Analysis.