China’s Embodied AI Boom: $13B Push to Outpace the US in Robotics

Chinese robotics firms are pouring capital into embodied AI, with funding totals rising sharply in early 2026. The trend marks a pivot away from showcasing humanoid robots performing tricks toward building AI systems that enable machines to learn, adapt, and operate in dynamic environments without extensive pre-programming.

The investment wave highlights growing competition between the US and China. While American startups like Generalist and Skild AI are advancing in rapid skill acquisition—such as training robots on tasks from short video clips—Chinese companies are leveraging their manufacturing expertise and access to real-world operational data to challenge US leadership. Industry observers suggest China’s strongest players are now competitive, though US firms still set the pace in how quickly robots can adopt and reuse new abilities.

The funding is reshaping the industry’s landscape. Meanwhile, established manufacturers like UBTech are aggressively recruiting top AI talent, with one executive role offering compensation of up to 124 million yuan ($18 million) to attract expertise in AI model development.

Beyond capital, Chinese firms are exploring different approaches to train robots, which require physical interaction data rather than just text-based inputs. Some companies, like Spirit AI, are hiring workers to record industrial tasks, while others, such as Ant Group’s Robbyant unit, have trained models on vast datasets—including 160,000 hours of human-guided robot movements. A few, like DexForce, rely on simulations to accelerate deployment, though critics argue these environments can’t fully replicate real-world physics.

The US maintains advantages in AI research and access to high-performance chips, but China’s deep ties to manufacturers and industrial users give it an edge in practical applications. The goal, as one executive put it, is to move robots from rigid programming to systems that operate with greater autonomy. Still, challenges remain: US export controls are restricting China’s access to advanced Nvidia chips, creating a bottleneck for training high-performance AI models.

This funding boom follows recent trends in the sector. Earlier this year, we reported on PsiBot’s $100 million round from industrial backers and DeepSeek’s $400 million investment in a Chinese robot AI startup—part of a broader expansion by AI companies into robotics. Meanwhile, European players are also scaling up, with a spinout from Mistral’s robotics unit seeking $231 million to compete.

Key questions moving forward: Can Chinese firms match or exceed US advancements in task generalization speed? How will chip restrictions force software optimizations—or push more development outside China? Will the $13 billion in funding lead to measurable gains in factory automation, where embodied AI’s potential is most evident?

The stakes are clear: whichever country leads in embodied AI could redefine industrial productivity. For now, the investment is flowing—but the race to practical breakthroughs is just beginning.

Sources: cio.economictimes.indiatimes.com