Capital expenditure on AI in China is expected to reach between 600 billion yuan and 700 billion yuan (US$84 billion to US$98 billion) in 2025, according to a report by Bank of America (BofA), as Beijing accelerates its push to rival the United States in a critical technology race.

The South China Morning Post reported that this projection would mark up to a 48% increase from 2024 levels, reflecting a sharp rise in enthusiasm for AI development across the mainland.

“Obviously China and the US are competing with each other,”

said Matty Zhao, co-head of China equity research at BofA Securities and head of Asia-Pacific basic materials, oil and gas research at BofA Global Research.

Government investment is anticipated to account for as much as 400 billion yuan of the total AI spending, the BofA report noted, while China’s major internet companies are expected to contribute up to 172 billion yuan.

The remainder will likely come from spending by major telecommunications network operators and allocations from special-purpose bonds.

China’s AI ambitions have been buoyed by the recent success of DeepSeek, a Hangzhou-based start-up that garnered global attention after releasing two open-source AI models, V3 and R1, developed with significantly lower costs and computing power than those typically required by Big Tech firms for large language model (LLM) projects.

Following DeepSeek’s achievements, Chinese tech giants such as Alibaba Group Holding and Tencent Holdings have pledged to ramp up their AI investments.

In February, Alibaba unveiled a 380 billion yuan investment plan for computing and AI infrastructure over the next three years, while Tencent’s AI-related capex for Q4 2024 surged nearly fourfold year-on-year to 36.6 billion yuan.

Zhao added that the central government has also stepped up efforts to build new AI data centres.

This is seen as China’s response to the United States’ Stargate Project, which is set to inject up to US$500 billion over four years into next-generation data centres across the US.

While the US is focusing heavily on IT hardware like semiconductors, Zhao noted that a larger share of China’s AI capital will be spent on constructing data centres and the necessary energy infrastructure.

“Energy resources represent an advantage for China,”

she said.

In line with this, Beijing recently released an action plan to coordinate data centre expansion with green energy infrastructure to support the massive power requirements of high-performance computing.

The BofA report further projects strong demand for resources: copper and power equipment are forecast to experience a compound annual growth rate (CAGR) of nearly 20% from 2024 to 2030, while demand for liquid cooling systems, essential for next-gen data centres, is expected to grow at a staggering 57% CAGR in 2025.

Despite facing US trade restrictions on advanced Nvidia processors and awaiting further breakthroughs from Huawei Technologies in chipmaking, China continues to rapidly expand its AI computing capabilities.

“Since a lot of the chip bottleneck depends on technology development, I think what the [Chinese] government can do is what they’re good at first,”

Zhao said.

 

Featured image credit: Edited by Fintech News Hong Kong, based on image by bluefish_ds via Freepik