US AI boom exposes critical mineral supply chain vulnerabilities to China dominance

As artificial intelligence accelerates the US data centre expansion, dependencies on foreign-critical minerals, mainly controlled by China, raise strategic concerns about domestic resilience and industrial security amid rapid technological growth.

The surge in artificial intelligence is accelerating an extraordinary build-out of data centres across the United States, and that boom is quietly exposing a deeper vulnerability: the supply chain behind the digital economy depends heavily on minerals that are often mined, processed and refined abroad. As companies race to add computing capacity for cloud services and AI models, the physical foundations of that expansion are increasingly shaped by access to critical materials rather than software alone.

A charted analysis by Visual Capitalist shows that the US is fully dependent on imports for several minerals central to semiconductor production, including arsenic, fluorspar, gallium, germanium, indium and tantalum. The country also relies on foreign supply for most of its platinum and more than a third of its palladium, both used in chip manufacturing. Beyond processors, data centre hardware also requires large quantities of copper, silver, tin and aluminium for wiring, soldering and cooling systems, while storage and magnetic components depend on rare earths and barite.

The strategic concern is not only the volume of imports but the concentration of control. According to a White House action issued in January 2026, the US was 100% net-import reliant for 12 critical minerals in 2024 and at least 50% dependent on foreign supply for 29 more. China remains the dominant producer and processor of many of these materials, giving it outsized influence over supply chains that underpin not just AI infrastructure but also national security and industrial policy.

That concentration has prompted growing calls in Washington and elsewhere for a more resilient minerals strategy. The Financial Times has reported that US officials are pushing for greater domestic production and processing, while the Centre for Strategic and International Studies has argued that building rare earth processing hubs would be essential to reducing exposure to China. Other analysts have suggested that AI itself could help speed the search for substitutes and new alloys, potentially shortening timelines for diversification. But for now, the AI race is being fought not only in chip design and model performance, but in mines, refineries and shipping lanes.

Source Reference Map

Inspired by headline at: ^[1]

Sources by paragraph:

Source: Fuse Wire Services