China’s AI push is prompting a closer look at two concerns: how far investors can carry humanoid-robotics valuations, and who may capture the economic gains from faster adoption. A September 26 analysis by the South China Morning Post framed both as potential risks raised by economists and industry leaders. It did not present a new policy, market outcome, or measured economy-wide result. Its value lies in identifying the pressure points now surrounding China’s AI-led economic transformation.
The market side of that discussion centered on Unitree Robotics. At the FutureChina Business Forum in Singapore, Daniel Zhang, FirstLight Capital’s managing partner and Alibaba’s former chairman and chief executive, said the pullback in certain Chinese humanoid-robotics names showed that earlier investor expectations had become overly high. Zhang’s remarks represent his interpretation of the market rather than a finding that every humanoid-robotics company is overvalued. He also described Unitree as a strong company led by a visionary entrepreneur, while arguing that no company could sustain expectations set at such elevated levels.
Unitree’s STAR Market Swing Sets the Valuation Context
Unitree’s path onto Shanghai’s technology-focused STAR Market shows why the valuation issue has become so visible. Material carried on the Shanghai Stock Exchange website on August 5, during the preliminary inquiry and before the final offer price was set, said Unitree planned to issue about 40.45 million new shares. That was equal to 10 percent of enlarged share capital of 404.5 million shares. The same exchange-hosted material said investment-bank estimates put a post-IPO valuation above 40 billion yuan, or US$5.9 billion, and described a preliminary issue-price estimate of around 104 yuan a share, subject to the inquiry process. Those figures were pre-pricing estimates, rather than a final trading valuation.
The eventual debut created a much sharper public benchmark. The BBC reported on August 19 that Unitree shares were offered at 150.80 yuan and finished their first trading day at 845 yuan, a gain of more than 460 percent from the offer price. That first-day surge was later documented by EastFrontier in Unitree Closes 460% Higher in Its Shanghai Debut. On September 26, SCMP reported that Unitree shares had fallen 55 percent from their peak less than a month after the listing. The two percentages use different starting points and dates: one measures the first-day close against the offer price, while the other measures a subsequent fall from the peak.
That distinction limits the conclusions that can fairly be drawn from Unitree’s trading. The 55 percent figure concerns a share-price move from a peak; it does not describe revenue, profitability, shipments, or a companywide operational loss. Likewise, the more-than-460-percent debut performance is historical context for the first trading day, not a measure that can be mechanically combined with the later decline. SCMP’s September 26 analysis therefore supplies a case study in volatile expectations around a prominent robotics listing, not proof of a settled valuation judgment across China’s entire humanoid sector.
Humanoid-Robot IPO Scrutiny Leaves Commercial Claims in Focus
The timing matters because Unitree’s debut has become a reference point for how investors read other planned robotics listings. An earlier SCMP report published September 15 said media reports had fueled speculation that mainland regulators could give greater scrutiny to humanoid-robotics IPO applicants, focusing on revenue-growth durability, earnings prospects, and innovation capabilities. The report described Deep Robotics and Leju Robot as among companies that had filed for IPOs while remaining unprofitable. Those reports concerned prospective oversight and sector conditions, rather than confirmation of a new rule or a completed regulatory action.
The new valuation analysis belongs beside EastFrontier’s earlier examination of China’s Humanoid IPO Slowdown Tests the Sector’s Commercial Claims. Together, the coverage captures a shift in the questions surrounding Chinese embodied AI: a high-profile market debut brought public attention, and subsequent reporting has focused on what scrutiny and investor expectations mean for future applicants. Zhang’s intervention adds a named industry voice to that conversation. It does not establish that his view is a consensus among investors, regulators, or robot makers.
Unitree’s pre-listing documentation also illustrates why precise labels matter. The August 5 exchange-hosted material described 40 billion yuan as an estimated post-IPO valuation and 104 yuan as a preliminary price estimate. The August 19 BBC account recorded an actual offer price of 150.80 yuan and a first-day close of 845 yuan. SCMP’s September 26 account then identified a 55 percent retreat from the peak. Keeping each number tied to its stated reference point avoids turning a sequence of market estimates and trading milestones into a single claim about the company’s current worth.
AI Adoption Could Deepen China’s Regional Divide
The second concern in SCMP’s September 26 analysis reaches beyond the stock market. A separate SCMP analysis published May 10 said large Chinese cities with deep pools of talent, capital, and innovative firms may be better placed to adopt AI than smaller cities and rural areas. Analysts and experts cited in that report warned that this gap in adoption could amplify regional inequality, at a time when wealth is already concentrated heavily along the coast. The article presented that as a prospective risk, not evidence that AI had already produced a confirmed widening in China’s wealth gap.
Lynn Song, ING’s chief economist for Greater China, told SCMP in May that AI gains would not be distributed equally and that people most directly connected to the core of China’s AI supply chain could benefit more. Her assessment gives the September discussion a concrete distributional lens: access to the people, capital, and innovative firms that support deployment may shape who participates first. It does not quantify income changes, establish an outcome for any province or rural area, or show that AI has caused losses for workers outside those networks.
That caution is especially important when robotics market enthusiasm and broader AI adoption are discussed in the same breath. Unitree’s post-listing price movement can illuminate the risks of elevated expectations around a single publicly traded humanoid-robot maker. The regional-adoption argument concerns a different question: whether technological capability and investment could remain concentrated in the cities and firms best equipped to deploy AI. SCMP’s analysis places both questions within China’s economic transformation, yet the available reporting does not establish a causal link between Unitree’s stock performance and regional inequality.
For policymakers, companies, and investors, the September 26 analysis offers a disciplined way to frame the next phase of China’s AI economy. The immediate facts are narrow: Zhang questioned excessive expectations after Unitree’s volatility, and analysts have warned that unequal access to AI’s inputs could leave regional gains uneven. The larger outcomes remain unresolved. Whether humanoid-robotics valuations settle on durable commercial evidence, and whether AI adoption spreads beyond the country’s strongest technology hubs, are questions the current reporting raises rather than answers.