Despite pouring hundreds of billions of dollars into artificial intelligence infrastructure and development, China’s AI boom is failing to deliver the broad economic lift that policymakers had hoped for, prompting analysts to question the structural bottlenecks hindering the technology’s impact. The disconnect between massive capital expenditure and sluggish national growth was a central theme at a media briefing in Beijing on Thursday, according to a report by the South China Morning Post.
“AI isn’t boosting China’s economy as much [as it is in the US], and we also have to worry about some of the negative side effects,” said Lu Ting, chief China economist at Nomura, during the briefing. According to Lu’s analysis, AI currently drives about half of the US economy, with investment in the sector growing roughly four times as fast as consumer spending.
In contrast, while China’s annual AI investment exceeds 1 trillion yuan (US$147.82 billion), its share of the overall economy is only about one-third of the US’s. “When you look at its overall contribution to GDP, it doesn’t really move the needle,” Lu noted.
The Hardware Bottleneck
A primary factor limiting the economic impact of China’s AI sector is restricted access to foundational hardware. US export controls have severely curtailed Chinese firms’ ability to procure the advanced semiconductors needed to train frontier models.
“The US has well-developed capital markets, so it’s easy for companies like OpenAI to raise money,” Lu explained. “And even if we, in China, want to invest in, for example, buying chips in bulk, we don’t have the means – [other countries] simply won’t sell to us. We’ve hit a major bottleneck here.”
This hardware constraint is starkly reflected in regional economic data. Taiwan, home to Taiwan Semiconductor Manufacturing Company (TSMC), reported a blistering 14.55 percent GDP growth in the first quarter of 2026, driven largely by global demand for AI chips. China’s Q1 GDP growth, meanwhile, stood at 5 percent. The World Bank recently revised its full-year growth forecast for China down to 4.2 percent.
“Without TSMC, and without the memory chips from South Korea’s Samsung and SK Hynix, it would be very difficult for US AI to advance into this era of large-scale investment,” Lu added, highlighting the integrated nature of the global AI supply chain from which China is increasingly isolated.
China’s own trade data illustrates the imbalance. In May, China’s integrated circuit (IC) exports surged 110.9 percent year-on-year by value to US$35.5 billion, but export volume grew by only 2.1 percent. Conversely, IC imports jumped 68 percent by value to US$56.6 billion, while import volume actually shrank by 1 percent, indicating that China is paying significantly more for fewer, likely less advanced, chips.
Uneven Spillovers and Displacement
Beyond hardware constraints, the economic benefits of China’s AI development are highly concentrated geographically and demographically. The industry is heavily clustered in top-tier hubs like Beijing, Shanghai, and Hangzhou.
This concentration threatens to exacerbate existing economic inequalities rather than alleviate them. Analysts warn of a bifurcation effect: as large language model providers in these top-tier cities deploy tools that automate lower-end white-collar work, the impact on smaller cities could be severe.
“When large language model providers in top-tier hubs begin replacing lower-end white-collar professionals, such as junior lawyers in small and medium-sized cities, the economic dynamic is no longer a positive outward spillover. Instead, it results in displacement,” Lu warned.
This dynamic is particularly concerning given China’s ongoing property crisis and sluggish domestic consumption. “When wealth and income are heavily concentrated among a small group of people in just a few cities, it’s very hard for domestic demand to drive the broader Chinese economy the way past booms used to,” Lu observed.
The failure of AI to serve as a broad-based economic engine presents a complex challenge for Beijing. While the government continues to elevate AI as core national infrastructure, the current trajectory suggests that without addressing fundamental hardware bottlenecks and ensuring more equitable distribution of the technology’s benefits, the AI boom may remain an isolated phenomenon rather than the catalyst for national economic rejuvenation.
The Long Game
Some economists caution against drawing premature conclusions from the current data. Transformative technologies historically take years, sometimes decades, to produce measurable macroeconomic effects. The commercialization of electricity and the internet both followed long gestation periods before their full economic impact became apparent in productivity statistics.
Moreover, China’s AI investment is still in its early stages relative to the scale of the economy. The 1 trillion yuan in annual AI investment, while substantial in absolute terms, represents a fraction of total fixed asset investment. As AI applications mature and diffuse across more sectors of the economy, from agriculture and logistics to healthcare and education, the macroeconomic impact is expected to broaden.
However, the structural challenges identified by Lu Ting are real and will not resolve themselves automatically. Closing the hardware gap requires sustained investment in domestic semiconductor development, a process that will take years even under the most optimistic scenarios. Ensuring that the benefits of AI diffuse beyond the top-tier coastal cities requires deliberate policy interventions, including targeted investment in digital infrastructure in lower-tier cities and retraining programs for workers displaced by automation.
For now, the gap between China’s AI ambitions and their economic reality serves as a sobering reminder that technological investment, however impressive in scale, is a necessary but not sufficient condition for broad-based prosperity.