China races ahead in humanoid robots

Images

×

German Chancellor Friedrich Merz and members of his delegation watch robots perform at a showroom of Unitree Robotics products in Hangzhou, eastern China, on Thursday. japantimes.co.jp

A solar farm in Nakai, Kanagawa Prefecture, in March 2016. Japan gets about a tenth of its electricity from solar panels despite having nearly no domestic production of photovoltaics (PVs). japantimes.co.jp

Sonic the Hedgehog, Castlevania's Alucard and the weak yet lovable Slime from Dragon Quest are just some of Japan's iconic gaming franchises celebrating midlife anniversaries in 2026. japantimes.co.jp

Japan agency aims for both decarbonization, growth japantimes.co.jp

techcrunch.com

‹ ›

Chinese humanoid robots have moved from gala-stage spectacle to factory and warehouse pilots, but Beijing’s push for automation is colliding with a labour market that is already fragile. Bloomberg reporting in The Japan Times describes President Xi Jinping’s dilemma: China wants the productivity and military-adjacent capabilities that come with leading in AI, yet the Communist Party also needs enough jobs to keep social stability.

The early market data points in one direction. TechCrunch reports that China’s leading humanoid makers—Unitree, Agibot, UBTech and others—are shipping more units and iterating faster than US rivals, helped by a hardware supply chain built up through the electric-vehicle boom. A China-based policy lead cited by TechCrunch says Unitree shipped roughly 36 times more units last year than US competitors Figure and Tesla, though the story notes the entire global market is still tiny—about 13,000 units in 2025 by one estimate, with uncertainty over how many were real commercial sales versus demos.

That small base matters because it shows where the immediate money is. In China, “operations-driven adoption” is being pulled forward by policy and procurement: local governments and state-linked firms can underwrite pilots that would be hard to justify on pure near-term return. That accelerates learning curves for manufacturers, but it also shifts the bill to institutions that are already carrying other social obligations.

The stress point is employment. Automation does not have to raise unemployment in the long run, but it concentrates disruption in the short run, and China’s adjustment mechanisms are constrained. Household consumption is weak, the private sector remains cautious, and many local governments are burdened by debt and property-market fallout. When a factory replaces workers with machines, the savings accrue to the firm and its backers; the costs—retraining, welfare support, and managing discontent—tend to land on local authorities.

Beijing’s answer so far is to treat robotics as both industrial policy and social policy. Officials sell automation as a response to labour shortages and as a route to higher-value manufacturing. The problem is timing: the regions most eager for investment are often the least able to absorb job losses, and the sectors easiest to automate—basic manufacturing, logistics, some service work—are precisely where large numbers of workers still earn their living.

There is also a measurement problem. Demo videos and staged performances can be broadcast nationally, while job displacement is dispersed and politically sensitive. TechCrunch notes that even shipment counts are hard to interpret at this stage, a fog that helps promoters and policymakers claim momentum without specifying trade-offs.

China’s humanoid industry is building scale faster than its rivals, and that advantage compounds. But the state cannot automate away the need for paychecks, and the institutions funding the robot rollout are often the same ones expected to keep a lid on the consequences.

In Hangzhou this week, visiting officials watched robots perform in a showroom, Bloomberg reported. Outside the showroom, the question is which budget line covers the workers no longer needed.