At the heart of the initiative is a mandate to source at least 80% of the technology—including AI chips—from local suppliers such as Huawei Technologies, effectively sidelining U.S. chipmakers like Nvidia and AMD from these state-funded projects . The plan also calls for state-owned enterprises, specifically China Mobile and China Telecom, to operate most of the data centers and ensure data interconnectivity, weaving scattered regional facilities into a cohesive “national computing network”
.
This AI infrastructure buildout is not a standalone project. It is the computing pillar of a much larger state-driven infrastructure program known as the “Six Networks,” a cornerstone of China’s 15th Five-Year Plan (2026–2030) that treats computing power as a strategic public utility .
The 2 trillion yuan AI data center plan represents a significant escalation in China’s state-capitalist approach to technology infrastructure. Key details reported by people familiar with the matter include:
The AI data center plan is the most prominent component of a vastly larger infrastructure initiative. In 2026, China’s Politburo elevated compute networks to the same strategic tier as water networks, power grids, communications, pipelines, and logistics—a concept formalized as the “Six Networks” (六张网) .
This move reframes AI computing from a tech industry concern to a “civilizational utility on par with drinking water and electricity” . The Six Networks include:
The scale of the overall Six Networks program is staggering. Total annual investment across all six categories is projected to exceed 7 trillion yuan (~$970 billion) in 2026 alone, a figure that has drawn comparisons to China’s 4 trillion yuan stimulus in 2009 . NDRC head Zheng Shanjie confirmed this estimate, describing it as a drive to “expand investment and stabilize growth” while supporting industrial development and national security
.
The user’s question references a potential total of 5 trillion yuan when power grid integration is included. This figure originates from a separate but deeply intertwined investment: a planned 5 trillion yuan ($722 billion) to upgrade China’s power grid over the same 15th Five-Year Plan period (2026–2030) .
This grid investment is not directly the “Six Networks computing pillar” budget, but it is essential for AI data centers to function. Goldman Sachs analysts have noted that China’s grid investment has evolved into a structural reshaping focused on supporting the power requirements of AI . The State Grid Corporation’s own plan to invest $574 billion through 2030 represents a 40% increase over the previous five-year period, with much of the focus on integrating renewable energy sources for power-hungry AI facilities
.
China is coordinating data center planning with energy infrastructure in resource-rich regions like Qinghai, Xinjiang, and Heilongjiang, aiming to power the AI boom with green electricity . Thus, while the 2 trillion yuan is specifically for computing infrastructure, the parallel 5 trillion yuan power grid upgrade is the necessary enabler, creating a total ecosystem investment of approximately 7 trillion yuan when considered together.
The plan crystallizes a fundamental strategic divergence. The U.S. model is driven by massive private-sector capital expenditure from a handful of tech giants. In contrast, China is building a state-led AI infrastructure directed and coordinated by central government planning agencies .
By mandating domestic chip sourcing, Beijing is also accelerating a technology ecosystem split. The policy reinforces a “two-loop” dynamic where China’s open-source AI strategy and domestic industrial policy mutually reinforce each other, aiming to build a self-sufficient computing supply chain from semiconductors to data center operations .
In summary, China’s reported 2 trillion yuan plan is a structured, state-directed push to build a national computing network, enforced by a hard domestic-sourcing mandate and operated by state-owned telecoms, all within the broader “Six Networks” strategic framework of the 15th Five-Year Plan .