
China’s artificial-intelligence rally has gone into reverse. Trillions of dollars have been wiped from Chinese stock markets in recent months amid rising anxiety about the future of AI, even as Beijing races the United States for dominance in the sector. The STAR 50 Index, which tracks technology and innovation companies on the Shanghai exchange, has plunged 40% since June, while other indexes of major companies have dropped about 20% over the same period. According to Bloomberg, the fall has erased $2.4tn from the Chinese market since May. Because AI-related stocks account for more than 40% of the stock market value in China and Hong Kong, according to Goldman Sachs, any dent in their value sends shockwaves across the whole domestic market. The sell-off raises a basic question: can a technology boom driven largely by the state sustain itself without strong consumer demand and open access to foreign markets?
The Chinese government has invested heavily in AI. This month a 60bn yuan state-backed AI investment fund started putting money into domestic developers. Beijing has also doubled down on robotics, also known as embodied AI, as the way to roll the technology out across society and as an answer to demographic problems such as an ageing population and a shrinking workforce.
The logic of the strategy is clear. A country with fewer workers needs machines that can do more, and a country competing with the United States wants its own technology stack. The difficulty is that stock prices ultimately depend on expected profits, and state support, however large, does not replace customers.
The most direct cause of the sell-off, according to the analysts quoted, is the lack of profits. Chinese consumers have proved stubbornly reluctant to spend in many parts of the economy, including technology, and many tech companies struggle to turn a profit. Dan Wang, China director at the Eurasia Group, summarised the problem: “This year, the profit margin is not as good as expected. And the US is raising rates. And China’s domestic demand is so weak.” She described a “general anxiety” among technology companies about whether the state-backed model will work in the face of weak consumer demand.
The point about US interest rates adds an international dimension. Higher rates in the United States make riskier, long-duration assets such as growth stocks less attractive and pull capital toward dollar assets. For Chinese technology companies whose valuations depend on distant future earnings, that is a particular headwind.
Robotics illustrates the gap between ambition and demand. Demand for robots is being driven by government agencies and research institutions, rather than by everyday consumers, for whom the products are still too expensive and not yet practical for home use. Humanoid robots and robotic dogs have been popular in the entertainment industry, but they are not yet useful in homes.The market has delivered its verdict on at least one prominent company. Shares in Unitree, China’s leading robot-maker, have halved since they began trading in August. Wang is blunt about the outlook: “It doesn’t look like embodied intelligence could turn into profits in the mass market.” The point is not that the technology lacks promise, but that a business needs buyers who can afford the product, and that sales to public institutions do not create the volume that justifies high valuations.
The second major cause is the outside world. Western countries, especially the United States, may extend bans on Chinese technology. The US has banned humanoid robots from China, citing national security risks, and is thought to be considering a ban on Chinese components in data centres. China is heavily reliant on exports, and restrictions hurt domestic companies that struggle to sell enough products at home. Weak domestic demand and restricted foreign markets therefore combine to squeeze the same companies from both sides. For investors, the risk is that the addressable market shrinks. A company that cannot sell abroad and cannot sell enough at home has a limited path to growth, no matter how advanced its technology.
The authorities want Chinese investors to put their money into domestic assets. In recent months China has restricted overseas investment, with more regulations taking effect on 1 December. According to the report, these moves have not led to greater investment in Chinese markets. This is a sign that capital controls alone cannot create confidence. If investors do not believe that returns will be adequate, they look for ways to move money elsewhere, and restricting those channels does not make domestic assets more attractive.
The picture is not uniformly negative. Wang notes that there is still optimism in other technology areas, such as AI research and satellites. The latter benefits from links to the military, which indicates long-term state support. Globally, the AI boom is expected to last for at least another year, which means that the sell-off may reflect problems specific to China, not a collapse of the sector as a whole.
The contrast with the United States is stark. On Monday, the American AI company Anthropic filed paperwork ahead of a planned IPO that could be among the biggest stock-market debuts of all time, with the company expected to raise as much as $100bn. China has no such blockbuster listings on the horizon. The comparison suggests that capital markets in the United States continue to provide generous financing to AI companies, while Chinese firms face a harder environment. Policymakers are under pressure to introduce stimulus measures to boost investment, but none are expected during what is described as a politically sensitive period. This week, top Communist Party officials gather for the fifth plenum, an important meeting that will lay the groundwork for next year’s 21st Party Congress, which is expected to give Xi Jinping a fourth term. There is no appetite for risky or potentially destabilising policy measures before then, even if they could help the economy.
Wang explains the incentive structure. Any stimulus measure would be “high risk,” she says: “What if it doesn’t work out? Then whoever came up with that policy, that guy would be held accountable.” In a system where accountability is personal and the political calendar is crucial, officials may prefer inaction to the risk of visible failure. That makes it less likely that markets will receive support in the near term.
Three conclusions follow. First, the sell-off reveals the limits of state-led financing. A 60bn yuan fund can support developers, but it cannot create consumer demand or lift profit margins by itself. Second, the strategy of making robotics a pillar of the AI effort faces a commercial test: if consumers do not buy, the sector depends on public procurement, which may not support current valuations. Third, the political calendar constrains the policy response, which means that the market adjustment may continue without a state backstop. For outside observers, the lesson is to separate technological progress from market valuation. China’s AI sector may remain strong in research and in areas favoured by the state, while many listed companies still face a profitability problem. For policymakers, the challenge is to find a path that raises household spending and eases external restrictions, neither of which is easy before the party congress.
The outcomes of the fifth plenum and any signals about economic policy, the performance of the STAR 50 and shares of companies such as Unitree, further US export or component restrictions, the effect of the new overseas-investment regulations from 1 December, and the pace of the 60bn yuan fund’s deployments will indicate whether the sell-off deepens or stabilises. The Anthropic IPO and other US listings will also affect global sentiment about AI valuations.
China’s AI sell-off is not simply a market correction. It reflects a structural tension between a state that wants to lead in technology and an economy where consumers are cautious and foreign markets are narrowing. Until profits improve, or policy changes after the political calendar clears, the gap between ambition and earnings is likely to remain visible in stock prices.






Comments