The AI Bubble Has Burst In China
- 27.10.2026, 14:50
The market overheating in China turned out to be significantly more severe than in the United States.
In China, the “bubble” in the artificial intelligence sector has already burst, according to Reuters. Shares in the STAR50 index, which includes market leaders in neural networks and semiconductors, have fallen by 40% over the past four months. The total market value has fallen by $480 billion.
One of the main reasons was the excessively high valuations of artificial intelligence-related companies. According to Goldman Sachs, these companies accounted for about 42% of the total market capitalization of Chinese A- and H-shares, as well as 57% of trading volume. Significant activity by retail investors contributed to the rise in stock prices but is now putting increased pressure on the market.
An additional warning sign for investors was the 20% decline in the CSI 300 index from its June high. As a result, many retail investors who bought shares in AI companies amid the hype have lost money. In addition, companies that have not yet gone public are now lowering their valuations.
Robot and microchip manufacturers have been particularly hard hit. For example, shares of the Chinese company Unitree, which produces humanoid robots, have lost about half their value since trading began in August 2026. Shares of chipmakers MetaX Integrated Circuits and Moore Threads Technology—which saw their stock prices surge by 700% and 400%, respectively, when they went public last year—have fallen by half since July.
“Like all other Chinese manufacturers, companies producing microchips and AI equipment will ultimately face a price war. There will be no exceptions. All bubbles burst sooner or later,” said Wang Zhao, a fund manager at Shanghai Zhuozhu Investment Management.
Reuters notes that these developments have dealt a serious blow to Beijing’s plans to develop a stable stock market and attract investment in high-tech industries.
A recent study showed that 95% of companies that invested in AI were unable to derive economic benefits from their investments. One of the challenges is the difficulty of retraining off-the-shelf AI tools for business tasks.
Many companies, having become disillusioned with AI, have begun rehiring employees they had previously laid off, suggesting that automation is not yet capable of fully replacing people to the extent that was expected. The study covered data on 2.4 million employees at 142 companies worldwide.