Digital dependency

Stable and abundant electricity is the fundamental resource for modern economies in the AI era

Total power use in China reached 10.37 trillion kilowatt-hours in 2025, crossing the 10 trillion kWh threshold for the first time and making China the world's largest consumer of electricity.

Against the backdrop of flourishing digital economy and the rapid rise of artificial intelligence, the significance of this milestone extends far beyond the energy sector.

Electricity is a unique strategic resource in the modern economy. Unlike primary fuels such as coal and oil, electricity, as a secondary energy carrier, cannot be stored in large quantities at low cost. Power generation, transmission and consumption therefore have to be kept in a real-time balance.

Therefore, operating a power system on the scale of 10 trillion kWh requires far more than large generation capacity. It also demands a highly complex grid network, sophisticated dispatch and coordination systems, and sustained infrastructure investment. In this sense, the power system is a manifestation of the country's industrial strength and capacity for long-term planning and systematic coordination.

Just as coal, steel and oil formed the bedrock of the industrial age, a stable and expansive power system has emerged as the foundation of the digital economy.

This is particularly evident in the development of the AI industry.

The AI race is often framed as a competition in chips and algorithms. In practice, computing power depends on something even more fundamental: electricity. Data centers and computing hubs are inherently energy-intensive facilities. A large-scale data center typically operates with power capacity exceeding 100 megawatts, and some supercomputing centers consume several billion kWh of electricity each year. Training large AI models often requires tens of thousands of GPUs to run continuously for weeks or even months, all sustained by enormous amounts of electricity. The International Energy Agency projects that electricity consumption from data centers will roughly double from 485 terawatt-hours in 2025 to 950 TWh in 2030, a level close to Japan's current total electricity consumption. These figures show that the AI industry is not merely a digital industry. It is also reshaping the structure of global energy demand.

The larger the scale of computing power, the higher the demands on the reliability and capacity of the power system. As the rapid rise of AI drives surging demand for computing power, the energy system is increasingly emerging as a core foundation underpinning the growth of the sector.

The scale and reliability of China's power system provide uniquely favorable conditions for the AI industry. China has the world's largest power generation system and transmission network, and it has accumulated extensive expertise in ultra-high-voltage transmission, renewable energy integration and cross-regional power dispatch. For the AI industry, these capabilities matter far beyond simply ensuring an adequate electricity supply. They are critical to the reliable operation of computing centers, the continued expansion of industrial clusters and the ability to keep energy costs under control.

A good case in point is China's "East Data, West Computing" project, which aims to align computing infrastructure with the geographic distribution of energy resources.

China's eastern regions have more developed digital economies and concentrated demands for data services, but face high electricity costs. By contrast, the western regions boast abundant wind, solar and hydropower resources, making the costs much lower.

Some computing tasks that are less sensitive to latency have been shifted to computing hubs in the western region of China. This has gradually created a more rational spatial division between data generation and computing infrastructure. The eastern regions mainly generate data and application demand, while some back-end computing and data-processing tasks can be handled by computing centers in the West.

By turning rich energy resources into strong computing power, the model ensures stable, low-cost energy to support the development of AI.

Energy researchers often talk about the "energy trilemma", which refers to the challenge of balancing efficiency, sustainability and security. The rise of AI gives this old challenge a new dimension: as the digital economy increasingly depends on computing power, the broader economy will be more dependent on a stable, clean and secure electricity system.

First, electricity is the direct energy input for computing infrastructure. Data centers, servers and cooling systems all depend on a continuous and reliable power supply. A more efficient power system therefore directly affects the cost and stability of AI development.

Second, electricity is easier to decarbonize than many forms of direct fossil fuel use. As the generation mix shifts toward cleaner sources such as wind, solar and hydropower, the expansion of computing power does not necessarily have to mean a proportional increase in carbon emissions.

Third, a stronger power system can also enhance energy security. As renewable energy capacity grows and more economic activity is supported by domestically generated electricity, dependence on imported fossil fuels can gradually decline, strengthening the resilience of China's energy system.

Therefore, over the long term, increasing the share of electricity in final energy consumption is not only an important pathway for green development, but also a more efficient and secure way of organizing the energy system.

Of course, operating the world's largest power system takes more than just a massive economy. Power infrastructure projects demand huge upfront investment, take a long time to build and deliver stable but slow returns. In many countries, such projects are often challenged by inadequate funding, grid expansion bottlenecks and the complexity of cross-regional coordination.

In China, decades of continuous power infrastructure expansion have been enabled by long-term planning and system-wide implementation. From the west-to-east power transmission program and ultra-high-voltage transmission corridors, to the construction of large renewable energy bases and national computing hubs, energy infrastructure development has always been guided by national-level planning and coordinated alongside industrial demand.

Reaching the 10 trillion kWh milestone is more than just an energy indicator. It signals the foundational capability required for the digital era. It demonstrates that China has both a massive industrial system and consumer market, as well as the energy bedrock needed to sustain the growth of AI, advanced manufacturing and the digital economy.

As AI development accelerates, a stable, green and efficient power system is vital to both the sustained expansion of computing infrastructure and the long-term trajectory of emerging industries. China's massive and evolving power system provides robust support for expanding AI application scenarios, upgrading industrial structures and advancing green development.

Looking ahead, as renewable energy, smart grids, energy storage technologies and computing infrastructure further integrate, the power system is expected to play an essential role in driving the mutually reinforcing digital and green transitions.

The author is an associate professor at the School of Applied Economics and the director of the Center for Research on Global Energy Strategy at Renmin University of China.

The author contributed this article to China Watch, a think tank powered by China Daily. The views do not necessarily reflect those of China Daily.

Contact the editor at editor@chinawatch.cn.