After years of perfecting humanoid robots' physical skills, Chinese firms are now channeling massive investments into AI software that enables robots to learn, adapt, and operate autonomously, marking a critical evolution in robotics capabilities.
- Chinese embodied AI startups raised $13 billion in H1 2026, five times the previous year.
- Focus is shifting from robot hardware to AI-driven adaptability and learning.
- US-China competition intensifies over the future of robot intelligence.
What happened
Chinese robotics companies have transitioned from focusing solely on physical capabilities—such as running, dancing, and complex maneuvers—to developing advanced AI software that enables robots to understand and adapt to their surroundings. This shift has sparked an investment boom in embodied AI, with funding reaching 93.5 billion yuan ($13 billion) across more than 300 deals in only the first six months of 2026, representing a fivefold increase from the previous year.
The investment surge is extending beyond robot manufacturers to specialist AI startups developing software that allows robots to learn new tasks from limited demonstrations and adapt without explicit programming. Leading Chinese firms such as X Square, Galaxea, and Spirit AI have reached valuations around $3 billion, highlighting the rapid rise of this new sector focused on robot intelligence.
Why it matters
This strategic pivot toward ‘robot brains’ addresses the critical bottleneck in robotics—intelligence. Robots capable of learning and adapting autonomously have far wider applications across industries such as manufacturing, logistics, and services. The ability to blend AI with mechanical skill could dramatically boost industrial productivity and reshape the global robotics market.
The race also marks an important front in the ongoing US-China technology rivalry. While US companies lead in some advanced robot AI technologies, Chinese firms leverage their manufacturing scale, access to extensive real-world training data, and close collaboration with industrial customers to compete aggressively. Winning this battle could confer significant economic and strategic advantages in automation leadership.
What to watch next
Attention will focus on the progress of AI models that allow robots to acquire and switch between tasks after brief demonstrations, as well as the methods used to train these systems using real-world data and simulations. Chinese startups like X Square and Spirit AI are investing heavily in collecting large datasets of physical interactions, aiming to close the gap with US competitors known for rapid skill acquisition in robotics.
Established manufacturers such as UBTech and Unitree are also deepening investments in AI research, signaling that the integration of embodied AI will increasingly drive the robotics industry's future. Observers should monitor funding patterns, technological breakthroughs, and international collaborations or restrictions that will shape the embodiment of intelligence in robotics.