Unitree Robotics shares have surged approximately two-thirds since October 15, closing at 672 yuan on Monday and elevating the Chinese robot maker's market value to 271.8 billion yuan ($40.8 billion). The rally reflects investor optimism that advancements in large language models will enable physical AI, allowing robots to perform complex tasks without extensive custom training. Founder Wang Xingxing stated at the World Physical AI Conference in Hangzhou on October 22 that breakthroughs in this area could occur within one to three years, citing OpenAI’s GPT-6 Astra as capable of directing robots with zero training.
JPMorgan upgraded Unitree to overweight from underweight on October 26, raising its price target to 995 yuan ($149.17). The bank revised its stance after engaging with executives and reviewing the company's humanoid lineup, moving away from previous concerns about an over-reliance on hardware relative to software capabilities. Unitree listed on the Shanghai Stock Exchange's STAR Market on August 19. Its G1 humanoid robot costs roughly $16,000, positioning it competitively against rivals, although researchers have flagged security flaws involving data streaming to external servers without user notification.
The stock movement highlights a broader market bet that chatbot technology can soon operate machines in the physical world. While Tesla has only begun producing Optimus robots with expected outside deliveries around the end of 2027, and Figure AI partnered with OpenAI in 2024, Unitree already sells its products. The company's valuation increase underscores the perceived potential for integrated AI to transform robotics from laboratory demonstrations into commercially viable consumer goods.
Unitree’s rapid valuation growth illustrates how capital markets are pricing in the convergence of generative AI and robotics hardware before widespread commercial deployment proves the technology’s reliability. Investors are effectively betting that the software layer—specifically large models capable of zero-shot instruction following—will solve the long-standing bottleneck of programming physical actions. This shift moves the competitive advantage from mechanical precision alone to the integration of sophisticated AI brains, potentially redefining which companies capture value in the humanoid sector.
However, this enthusiasm contrasts sharply with unresolved operational risks, particularly regarding data privacy and security architecture. The presence of flagged vulnerabilities where audio and spatial data stream to external servers without user consent introduces significant regulatory and reputational hazards. As JPMorgan and other institutions upgrade their positions based on executive promises rather than verified mass-market adoption, the sustainability of Unitree’s premium depends on whether it can secure its software stack while scaling production. Failure to address these security gaps could invite stricter oversight or consumer backlash, undermining the very trust required for household robotic assistants.
