
Beijing's $295 billion AI buildout would require 80 percent domestic chips, locking out US suppliers
Quick Answer
China is set to invest $295 billion in an AI data center network over five years, requiring 80% of technology from domestic suppliers like Huawei, effectively excluding US suppliers.
Quick Take
Taiwan is also considering criminalizing AI chip smuggling to China.
Key Points
- Investment of $295 billion aimed at nationwide AI infrastructure.
- 80% of technology sourced from domestic suppliers, primarily Huawei.
- US suppliers will be largely excluded from this initiative.
- Taiwan may criminalize AI chip smuggling to China for the first time.
📖 Reader Mode
~3 min readBeijing's $295 billion AI buildout would require 80 percent domestic chips, locking out US suppliers

Midjourney prompted by THE DECODER
Beijing is preparing a massive investment program to expand AI infrastructure nationwide. Meanwhile, Taiwan is considering making the export of AI chips to China a criminal offense for the first time.
China plans to invest roughly 2 trillion yuan ($295 billion) over the next five years to build a nationwide data center network, according to Bloomberg. The National Development and Reform Commission is drafting a blueprint that lays out a web of interconnected computing hubs.
State-owned companies like China Mobile and China Telecom would operate most of the data centers. At least 80 percent of the technology used - including AI chips - would come from domestic suppliers like Huawei. That would effectively shut out Nvidia and AMD.
Funding would come mainly from ultra-long-term government bonds and state investment funds, supplemented by bank loans and private capital. The plan is part of the "Six Networks" program announced earlier this year, which covers critical infrastructure from water and power to computing capacity. Under the plan, China's AI industry would grow to more than 10 trillion yuan in total volume.
China's most aggressive AI push yet
The investment target is Beijing's most ambitious effort yet to lay the groundwork for Chinese AI development, Bloomberg reports. By 2028, scattered data centers would be linked into a single coherent network. Including power infrastructure, the total investment could reach at least 5 trillion yuan.
FUS tech giants like Meta and Microsoft are planning about $725 billion in AI spending for 2026 alone. Chinese data centers are cheaper to build, though, thanks to lower labor and construction costs. And the 2 trillion yuan figure doesn't include private spending from companies like Alibaba and Tencent.
Nine Chinese AI chips - from Huawei, Alibaba, Shanghai Biren Technology, and others - recently passed a government security review, Bloomberg reports. That clears the way for their use in security-sensitive sectors. According to Bloomberg Intelligence, the main beneficiary of the plan is the broader economy, not individual private companies. Huawei stands to gain the most.
Taiwan considers criminalizing AI chip smuggling to China
As China ramps up its buildout, Taiwan is weighing much stricter export controls on AI chips headed to China, Bloomberg reports. Right now, unauthorized export of such chips from Taiwan to China isn't a criminal offense. Authorities can only go after suspects on other charges, like document forgery.
As part of ongoing trade talks with the US, Taipei is now looking at restricting sales to all Chinese customers above a certain computing power threshold and not just companies on an export blacklist like Huawei. That would let Taiwan prosecute AI chip smuggling to China as a crime for the first time.
In May, Taiwanese authorities arrested suspected chip smugglers for the first time, but could only charge them with document fraud. Investigators suspect Nvidia servers were rerouted through Japan to Hong Kong.
— Originally published at the-decoder.com
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from The Decoder
See more →
An AI model programmed nonstop for 19 days on a single MirrorCode task that cost $2,600 to run
Epoch AI's MirrorCode benchmark reveals Claude Opus 4.7 as the leader with a 56% solve rate, reconstructing a 16,000-line toolkit in 14 hours. Despite this, all models tested struggle with the most complex tasks, highlighting limitations in current AI capabilities. The single task consumed $2,600 over 19 days, raising questions about cost-effectiveness in AI development.

