China is accelerating ahead in the global artificial intelligence (AI) supercycle, with enterprise adoption, industrial deployment, and ecosystem expansion converging to widen its lead over other major markets.
The AI market has entered a new phase. Global enterprise AI spending is projected to reach US$940 billion in 2026, rising to US$2.1 trillion by 2029, according to IDC. The shift marks a transition from foundational infrastructure to large-scale enterprise application.
“The global AI industry has entered a super cycle, and the market is now moving from infrastructure build-out to enterprise application explosion,” said Kitty Fok, managing director at IDC China.
China is moving faster than most. “We’ve never seen the pace of change move faster than it is right now,” said Lorenzo Larini, CEO of IDC. “China is not a market you can afford to observe from a distance because it is a technology force actively shaping how the world moves.”
Robotics and embodied intelligence surge
China’s leadership is particularly evident in robotics and embodied intelligence. IDC forecasts spending in this segment will surge from US$1.4 billion to US$77 billion within five years, representing a 94% compound annual growth rate. The country is also on track to become the world’s largest robotics market by 2029, with domestic manufacturers already leading global shipments across multiple categories.
This growth aligns with broader global trends. According to the International Federation of Robotics, China already accounts for more than half of global industrial robot installations, underscoring its scale advantage.
Tokens, agents and the economics of AI
The economics of AI are also shifting. Tokens—the unit of AI output—have become the key determinant of both cost and value.
Steven Carlini, vice president, innovation and data center at Schneider Electric, believes it is important to understand what tokens are in the context of AI.
"They are how people pay for AI working models, also known as reasoning or inference models. In working AI models, tokens are used for input queries as well as output intelligence of prediction, content generation and reasoning. Users can pay based on token use." Steven Carlini
China’s Model-as-a-Service (MaaS) market is forecast to grow at a 1,154.9% CAGR between 2024 and 2030, reaching 40,000 trillion token calls in 2026 and generating RMB18.6 billion in revenue.
More than 60% of leading Chinese enterprises have already embedded generative AI into core operations, reflecting a move from experimentation to execution.
As Zhenshan Zhong, vice president at IDC China, noted, enterprises are transitioning from “generation” to “execution”, with agents driving real business outcomes.
Efficiency becomes the battleground
Performance metrics are evolving alongside adoption. Traditional measures such as floating-point operations per second are giving way to “tokens per watt”, reflecting the efficiency of AI output relative to energy consumption.
IDC expects inference workloads to account for over 70% of computing demand by 2027, with edge infrastructure outpacing centralised data centres.
This shift is mirrored globally. Gartner forecasts that by 2026, more than 80% of enterprises will have used generative AI APIs or deployed AI-enabled applications in production environments.
Policy and industrial transformation
China’s 15th Five-Year Plan, launched in 2026, is reinforcing this trajectory through a focus on digital sovereignty, ecosystem expansion, and capability exports. Industrial AI is already transforming manufacturing, integrating intelligence across supply chains and operations.
At the same time, AI-native devices are redefining consumer and enterprise markets. While shipments are stabilising at around 900 million units in 2026, value is shifting from hardware to intelligent services and ecosystem integration.
