The race is on.
China is aggressively positioning itself to dominate the emerging humanoid robotics industry, outpacing the United States through massive government support, a deep domestic supply chain, and rapid commercialization. The Wall Street Journal highlights how Beijing views “embodied AI” (meaning AI integrated into physical systems like robots) as a strategic priority for the next five years, mirroring its successful playbook in electric vehicles (EVs).
Over 140 humanoid-robotics companies have sprung up across cities like Shenzhen, Suzhou, Beijing, and Hangzhou. These firms benefit from a vast ecosystem of local suppliers providing components such as sensors, batteries, motors, and joints at low cost and with quick turnaround. This enables rapid iteration and innovation. For instance, UniX AI in Suzhou can source about 80% of its parts within a one-hour radius, allowing immediate problem-solving and design tweaks. Companies like UniX AI demonstrate practical applications: its wheeled humanoid Panther adjusts bedsheets, picks up trash, and handles laundry in hotels, with units priced around $12,600 and hundreds already deployed in China.
Beijing and local governments are fueling this boom with substantial incentives. That’s what China does. Since late 2024, cities including Beijing and Shenzhen have established investment funds totaling over $26 billion (per Morgan Stanley estimates) to inject capital into the sector. Support includes discounted land and office rents, favorable loans, subsidies covering about 10% of robot purchase prices for buyers, and free or reduced-cost facilities for startups. State-owned enterprises and government agencies act as early adopters, deploying robots in museums, events, airports (e.g., luggage handling), factories, and even as “robocops” for traffic control. These real-world uses generate valuable data to improve performance and build market demand.Chinese firms are scaling production quickly. In the second half of 2025, orders exceeded $300 million, with companies like Shenzhen-based UBTech selling to international clients such as Texas Instruments and Airbus. Morgan Stanley forecasts up to 100,000 humanoid units shipped in 2026, with adoption accelerating faster in China than in the U.S. Some observers, including a Singapore industrial park representative, note that Chinese products are often the only viable option available.
Elon Musk has repeatedly warned of China’s threat, calling it “an ass-kicker, next level” and acknowledging no significant competitors outside China. Tesla’s Optimus robot relies on Chinese suppliers for key components like roller screws and hand motors for mass production, raising U.S. concerns about supply-chain dependence. The U.S. leads in foundational AI models (from Nvidia, Google, etc.), powering robot “brains,” but China’s hardware advantages and manufacturing speed give it an edge in deployment.
Skeptics, however, warn of a potential bubble. The industry remains nascent, with technical challenges in dexterity, locomotion, and reliability. Hype events—like a 13-mile humanoid marathon where robots needed human help or failed—underscore limitations. Rapid proliferation risks overcapacity, price wars, and unprofitability, similar to China’s EV sector where hundreds of brands competed fiercely. To mitigate this, China’s government is drafting technical standards, forming a standards committee, and tightening oversight on public listings to weed out weak players and promote sustainable growth.Overall, China’s strategy combines state mobilization, supply-chain depth, talent (including returnees from U.S. universities), and aggressive commercialization to challenge U.S. leadership. While the U.S. responds with policy efforts like a potential executive order to boost domestic robotics, China’s momentum could make it the first to achieve widespread humanoid deployment—if the sector avoids bursting.
We’re seeing plenty of companies in the generative AI space get ahead of customer demand by rolling out AI practically everywhere, and often overselling capabilities. We can probably expect something similar with robots. When will they be good enough that people and businesses will actually want them?
