China Launches Xingshu Plan AI Satellite Network Utilizing Gallium Technology

The Xingshu Plan and In-Orbit Computing

On July 19, 2026, China officially initiated the Xingshu Plan, a commercial space-based computing network led by Shanghai Xingshu Tiansuan Space Technology. The project aims to deploy a fleet of up to 1,000 AI-enabled satellites designed to process data directly in orbit. This network is intended to resolve the limitations of ground-based data centers by reducing latency and minimizing the need for heavy bandwidth during data transmission.

The strategic intent behind this deployment was highlighted by President Xi Jinping, who described the effort as a fundamental step toward building a new global AI order. The initiative represents a transition from reliance on traditional data downlinks to immediate, in-orbit information downlinks, allowing for faster processing cycles that effectively bypass the constraints of terrestrial infrastructure.

Gallium Dominance and Surveillance Capabilities

The effectiveness of China’s space-based assets is tied to its control over specific raw materials. China’s surveillance satellites, which operate from a geosynchronous orbit 36,000km above Earth, rely on gallium nitride (GaN). This material is essential for active electronically scanned array radars and electronic warfare systems.

China maintains a dominant position in the supply chain for this metal, producing more than 95 per cent of the world’s primary gallium. Because the metal is primarily extracted as a by-product of alumina production, China’s status as the world’s largest aluminium producer has provided it with an effective monopoly. Since the launch of its high-orbit surveillance satellite in August 2023, China has implemented restrictions on gallium exports, a move that experts link to broader trade tensions regarding semiconductor technology.

Global Compute Competition and Ground-Level Resistance

As China expands its orbital infrastructure, Western technology firms are adjusting their internal strategies to compete in what is being termed a compute war. OpenAI has recruited Uday Ruddaraju, formerly of xAI and AWS, to serve as its new CTO of Compute, with a mandate to develop a massive infrastructure footprint capable of supporting future frontier models.

However, the rapid adoption of AI-driven automation is facing significant pushback from labor forces on the ground. In South Korea, over 40,000 unionized Hyundai Motor workers recently staged the auto industry’s first-ever strike triggered by the introduction of humanoid robots. The workers are demanding Atlas consent regarding the deployment of Boston Dynamics robots on factory assembly lines, citing fears of job displacement.

The volatility of these automated systems was further underscored in San Francisco, where a power outage at PG&E forced a temporary suspension of Waymo’s robotaxi service. These incidents highlight the ongoing tension between the push for advanced AI efficiency and the necessity of maintaining stability in existing human and utility systems.

Strategic Stakes in Surveillance and AI

The technological gap between China and the United States remains a focal point of the current geopolitical climate. The high-orbit satellites currently utilized by China provide surveillance of a third of the Earth’s surface. This capability allows for the monitoring of large geographic areas and moving targets, an objective that the United States has pursued for decades without achieving similar success.

While the Xingshu Plan promises to redefine data processing through its AI-enabled network, the reliance on restricted minerals such as gallium suggests that the future of space-based computing will remain inextricably linked to global trade policy. As of late July 2026, the primary uncertainty remains how Western nations will respond to the dual challenge of China’s orbital dominance and the growing domestic pressure to protect labor markets from rapid AI integration.

Sources: SCMP, brownstoneworldwide.com.