China Flexes AI Muscle with Constellation of 2,800 Satellites

With the launch of 12 satellites last month, has China taken an important leap ahead of other countries in the race to lead in AI? Perhaps. Beijing’s campaign to dominate AI is systemic, state-backed, and increasingly orbital, a new report from the Special Competitive Studies Project and Strider Technologies suggests. And that could put the U.S. and other Western countries at risk of falling prey to industrial espionage.

It’s clear from the China’s AI Infrastructure Surge study that China has made great leaps in its quest to dominate in AI. The more than 250 advanced AI-focused data centers that China already has built or are under construction across the country aren’t just industrial server farms. They’re strategically positioned nodes in a national computing architecture designed to fuel both economic power and military modernization.

But the real shift—and perhaps the most geopolitically significant one—isn’t happening on Earth, but rather in Space.

In mid-May, Chinese startup ADA Space, in collaboration with Zhejiang Lab, launched the first 12 satellites of a planned network of 2,800. Each satellite is designed to perform real-time AI processing while in orbit. This constellation, interconnected by high-speed laser links, represents a prototype for something profoundly disruptive: the ability to process and act on data in space, independent of ground infrastructure.

And from there it won’t be too long before the debut of surveillance satellites that can analyze threats in real-time and disaster-monitoring systems that trigger alerts without waiting for ground confirmation. In military terms, it’s the genesis of orbital decision-making.

Those orbital capabilities are backed by China’s growing domestic computing base—forming a vertically integrated, end-to-end system of terrestrial and space-based AI infrastructure. “This isn’t opportunistic,” said Christopher Gragg, a Strider analyst. “It’s strategic. Every element is coordinated.”

China’s model differs from that of the West. In the U.S., AI innovation often flows from the bottom up—driven by academia and private sector investment. In China, the execution strategy is decidedly top-down, aligning AI infrastructure with national industrial policy, military doctrine, and long-term geopolitical positioning.

Former Google CEO Eric Schmidt has openly contended that “the edge” in AI is no longer a server but rather an orbital platform. And China has clearly gotten there first.

Washington is countering China’s AI ascent with a familiar, and increasingly insufficient, toolkit: restrict access to cutting-edge U.S. technology. That includes blocking Nvidia’s high-performance GPUs, curbing the export of advanced lithography equipment essential for chip manufacturing, and tightening controls on Chinese students and researchers in sensitive fields.

But that’s not enough. The problem isn’t the measures themselves—it’s the underlying assumption that chokepoints still exist in a truly globalized, digitized economy. China has treated each restriction as a design challenge, accelerating investment in its domestic semiconductor ecosystem and building state-backed fabs and packaging facilities while subsidizing the development of domestic alternatives to Western chips. Companies like Huawei, SMIC, and Biren Technology are all working—under various levels of international scrutiny—to reduce dependence on U.S. components, particularly prudent as President Trump uses tariffs as a bludgeon to bring chip manufacturing back to U.S. shores.

But technical substitution is only one dimension. As Greg Levesque, CEO of Strider Technologies, explains, “This isn’t about blocking individual transactions anymore. It’s about tracking entire ecosystems.” That includes a complex web of shell companies used to re-route restricted components, joint ventures designed to obscure ownership, and gray-market procurement channels that stretch across continents. Supply chains are no longer linear—they’re adaptive, encrypted, and sometimes state-augmented.

Meanwhile, China’s military-civil fusion policy continues to erode the boundary between the commercial and the strategic. The country often builds infrastructure projects that appear purely industrial—such as deep-sea or inland data centers or even smart city networks—with dual-purpose objectives. A satellite capable of edge computing that in isolation may appear to be an academic experiment when deployed as part of a 2,800-satellite constellation integrated with terrestrial AI clusters becomes a weaponized infrastructure layer capable of delivering real-time, sovereign control over information and surveillance domains.

This layered approach is what makes China’s AI surge difficult to counter because it doesn’t rest on a single technology or firm but rather a globally distributed computing network that serves as both an economic engine and a geopolitical instrument. Every satellite, every data center, every chip fab is a brick in a broader system designed to give Beijing strategic autonomy—not just from the U.S., but from the current global tech order.

As the West forges ahead with their own plans to dominate the AI landscape, the Strider report recommends that industry leaders and policymakers take steps now to gain traction and guard against inevitable and formidable espionage campaigns. The researchers urge policymakers to restrict those developers of security-relevant AI infrastructure, leverage PRC dependence on foreign software–still a clear weakness–to apply strategic pressure, and bolster collaboration internationally to monitor and encounter PRC’s efforts to expand AI infrastructure. 

Beyond that, industry leaders should monitor China’s AI activities, become intimate with the competitive landscape, get rid of the vectors for technology transfer, and just generally raise awareness among top talent of China’s tactics.