China’s AI industry is primarily plagued by a critical disconnect: policy and capital injections are vastly outpacing actual market demand. Policy mandates are prematurely pushing immature technologies to market, generating ineffective supply.
Driven by state policy and capital, China’s AI sector is rapidly expanding. Official data shows the core AI industry scale surpassed RMB 1.2 trillion (approx. NT$5.6 trillion) in 2025. In August 2025, the State Council launched the “AI Plus” initiative, deepening deployment in key sectors like smart manufacturing, healthcare, and transportation. The Ministry of Industry and Information Technology (MIIT) projects China’s AI-related market will reach US$1.4 trillion (approx. NT$44 trillion) by 2030, becoming the primary economic engine during the 15th Five-Year Plan period (2026–2030).
However, Meng-Chun Liu, Director of the China Economic Research Institute at the Chung-Hua Institution for Economic Research (CIER), warned that supply-side expansion in certain AI segments is already outpacing demand. Without synchronized commercialization, a robust domestic market, and practical application scenarios, these artificially inflated sectors face severe challenges such as capital idling and the risk of a market bubble.
Upstream Hardware Far Outpaces Downstream Applications
China’s AI growth is starkly bifurcated: upstream hardware and core technologies are growing significantly faster than application endpoints. AI chips and servers serve as the primary growth engines. The China Commercial Industry Research Institute forecasts the domestic AI chip market will reach RMB 381.39 billion in 2026. Factoring in IDC’s estimate that domestic chips hold a 41% market share, China’s homegrown AI chip market will be worth approximately RMB 156.4 billion in 2026. Meanwhile, the AI server market is projected to hit RMB 285.9 billion to RMB 350 billion in 2026; compounding at a 35% annual growth rate, it could reach RMB 1.2 trillion to RMB 1.5 trillion by 2030.
Chronic Supply-Side Bias Leaves a Deficit of Buyers
Beneath these glowing figures, Director Liu identified a fundamental flaw: rampant policy and capital inputs are completely detached from actual demand. Immature technologies caught in policy-fueled hype cycles are rushed to market, creating ineffective supply. He stressed that the tech industry must remain grounded in practical applications and market demand, relying on domestic consumption to anchor the supply chain. China’s chronic reliance on supply-side policies—artificially stacking industrial scale through subsidies, tech parks, and government guidance funds—has led to a dearth of demand-side stimulus. Consequently, high-tech capacity is left without enough buyers, a vulnerability magnified by anemic consumer momentum.
Macroeconomic data reflects this imbalance. According to May 2026 National Bureau of Statistics data, high-tech manufacturing profits surged 44.8% year-over-year in the first four months of 2026. Profits for upstream AI sectors like electronic specialized materials, optical fibers, and optoelectronic devices soared 601.7%, 347.6%, and 51%, respectively. Conversely, consumer-linked sectors such as automobiles, specialized equipment, and furniture saw broad declines, underscoring that momentum is heavily concentrated upstream.
Fu-Kuo Chung, an analyst at CIER’s China Economic Research Institute, pointed out that China’s economy is caught in a transition between old and new growth engines. AI-driven industries are highly capital- and tech-intensive, offering limited employment absorption and hindering labor from smoothly transitioning into emerging sectors. He added that slowing household income and consumption growth mean AI-driven efficiency gains are failing to translate into consumption expansion, further stalling the domestic economic cycle.
Analyst Chung argued that while AI improves the efficiency of “making the pie,” it exacerbates income inequality and diminishes worker bargaining power. Data shows that in Q1 of 2026, the growth rates of both disposable income and consumer spending slowed, failing to outpace GDP growth. If the fiscal and tax mechanisms for “dividing the pie” are not concurrently optimized, this supply-demand imbalance could morph from a short-term issue into a chronic structural dilemma—leaving the AI industry with nominal output value but no virtuous economic cycle to sustain it.
Capital Idling: AI Startups Rely on Subsidies
Regarding funding structures, Director Liu pointed out that Chinese AI startups remain highly dependent on government subsidies and state-backed capital, leading to “capital idling.” Unlike mature markets such as the U.S., which rely on institutional investors for market screening, China leans heavily on official “guidance funds.” These investment decisions are often politically driven to showcase performance rather than being commercially sound, resulting in low capital efficiency.
Similar issues plague China’s computing infrastructure. Regional intelligent computing centers are expanding rapidly. While some municipalities can generate short-term output value by simply networking equipment and stacking systems to inflate on-paper computing power, they face immense operational pressure from energy consumption and maintenance costs. With AI applications still in their infancy and commercialization yet to fully unfold, the return on these massive computing investments remains unproven.
As for emerging applications, embodied AI and the low-altitude economy are heavily prioritized by local governments, yet their commercial viability is highly questionable. Analyst Chung observed that many humanoid robot demonstrations currently rely on pre-programmed movements within strictly controlled environments, rendering them essentially sophisticated automated systems. Choreographed walking and dancing are forms of automation, not AI robots capable of spatial comprehension and autonomous decision-making. A machine executing preset routines is closer to a music box than an AI agent, lacking the core advantages of the AI era. The key to industrializing humanoid robots lies in “autonomous” comprehension and decision-making in complex environments—a threshold current models still struggle to cross in the physical world.
Low-Altitude Economy “Lacks Feasibility”
Regarding the low-altitude economy, Director Liu argued it lacks fundamental market demand and physical feasibility in China. Unlike the U.S., which boasts sprawling, low-density residential areas and over 7,000 private airports, China’s population is highly concentrated in high-rise urban centers, and its airspace is under strict military and administrative control. Under these constraints, flight density and safety regulations make it nearly impossible to build a scalable commercial market, leaving this sector largely theoretical.
In extreme environments such as space, where human intervention is difficult, AI and robotics hold greater long-term potential.
China is pouring massive capital and resources into fully developing its AI industry, marking it as the absolute priority for the 15th Five-Year Plan. However, whether this AI boom represents a genuine new growth curve or a policy- and capital-inflated mirage ultimately depends on whether the demand side can actually absorb the expanding supply. Without solid demand, this seemingly glorious boom may inevitably burst into a bubble.
Source: United Daily News (June 22, 2026). China Bets Heavily on AI, Experts Warn of a Mirage. United Daily News. https://udn.com/news/story/124068/9579699