China’s AI Boom Is Reshaping Its Power Grid, Driving Green Energy and Grid Stability Initiatives

China’s rapid expansion of artificial intelligence (AI) is fundamentally transforming the nation’s power grid, prompting a strategic shift toward renewable energy and innovative grid management solutions. The surge in AI computing demand, which has seen daily token usage surpass 140 trillion by March 2026—a 1,000-fold increase from early 2024—is driving a corresponding rise in electricity consumption. Computing-related power use grew from 82.4 billion kWh in 2019 to 196 billion kWh in 2025, with projections reaching 500–700 billion kWh by 2030. This growth has sparked a stock rally for state-owned utility Datang International Power Generation Co. Ltd., whose shares surged over 130% in a month after launching a solar plant in a cloud computing hub, though analysts caution the market overreacted to the “computing and power synergy” concept as Datang had no direct operational synergy project.

The Chinese government has formally endorsed computing-power coordination, including it in the March 2026 work report. On May 8, 2026, four top bodies issued a joint action plan targeting a secure green energy system for AI by 2027 and full two-way empowerment by 2030. A key technical challenge, according to Wang Zesen of State Grid Jibei, is not total electricity volume but the grid’s ability to deliver stable instantaneous power to specific locations at specific times. This issue is exacerbated by AI servers causing instantaneous fluctuations up to 10,000 kW and the “dual-high” stability risk in regions with high renewable generation and large data center loads.

To address these challenges, China is implementing its “Eastern Data, Western Computing” strategy, launched in 2022, which directs data centers to resource-rich western hubs that must source over 80% of their power from green energy. Examples include a Datang solar farm and planned wind farm in Zhongwei, cutting electricity costs to 0.36 yuan/kWh—20% below the local average—and Ulanqab in Inner Mongolia, which uses a “source-grid-load-storage” model with 67% green power and low latency to Beijing. However, local grid strains are evident: Datong’s computing sector consumed over 6 billion kWh in 2025, a 40% jump surpassing coal, accounting for 26.2% of grid load. A severe timeline mismatch persists, as data centers can be built in 8–24 months while substations take 3–5 years, and opaque utilization rates (10–80%) complicate planning.

China is testing virtual power plants (VPPs) to manage load flexibility. In one test, Shanghai migrated inference tasks to Fujian in three minutes to shed 50 kW, and Guangdong integrated three telecom data centers into its spot market. Despite these advances, true synergy remains nascent due to data transmission latency, physical hard-drive transport, and high bandwidth costs. The path forward requires unifying fragmented power, computing, telecom, and carbon markets, with tech giants like Alibaba and Tencent investing in next-generation nuclear technology. Wang Yongzhen expects the experimental phase to last another three years, with the ultimate goal of integrating security, green energy, and economic efficiency. Analysts like Gao Xing of China Securities dismiss fears of nationwide shortages, noting AI load is only about 1% of total consumption, but warn of a fundamental shift in grid operation.