From a strategic perspective, China’s near‑monopoly in the processing and refining of critical minerals—from rare earths to battery materials—constitutes a structural vulnerability in the global supply chain, one that technology firms such as Alphabet Inc. can no longer safely ignore. The concentration of processing capacity, combined with the instrumentalization of export controls as a tool of statecraft, introduces a level of supply risk that directly threatens the hardware backbone of Alphabet’s data centers, consumer devices, and emerging artificial intelligence infrastructure. We would do well to understand these dependencies not as a temporary market distortion but as a durable feature of great power competition.
A Structural Imbalance in Critical Minerals
China’s dominance in rare earth processing is all but absolute. Current estimates place its share of global oxide separation capacity at 91.3% 16, with the broader processing figure hovering between 80% and 90% 4,15,20. The nation supplies some 90% of the world’s permanent magnets 14 and processes essentially all heavy rare earth elements 14, a category vital to high‑performance applications. This leaves the United States, which sources roughly 70% of its rare earths directly from China 14, acutely dependent—and lacking any domestic heavy rare earth refining capability 14. Similar patterns extend to an array of other technologically critical materials: China refines 99% of global gallium 12, 60% of germanium and 80% of antimony processing 14,16, 68% of nickel 5, 73% of cobalt 5, 59% of lithium 5, and over 90% of battery‑grade graphite 5. In the lithium‑iron‑phosphate battery segment, the concentration is even sharper: over 98% of LFP cells and 99% of LFP cathode active material originate in China 5. Such figures bespeak not merely scale but a deep structural advantage built through deliberate policy choices.
That advantage rests on a foundation of state‑directed industrial policy and favorable factor costs. Industrial electricity prices in China run lower 9, environmental oversight lighter 19, and the state’s ownership of land streamlines permitting and infrastructure 21. The country graduates over 500 rare earth specialists each year 25, guaranteeing a deep pool of technical talent. Domestic overcapacity in sectors such as steel 6,18 creates secondary market pressures that further reinforce cost competitiveness. These elements are reinforced by industrial policies that prioritize domestic manufacturing 6, restrict foreign investment and outward technology transfer 22, and increasingly enforce localization under an “in‑China‑for‑China” framework 11, compelling multinationals to shift research and development activity onto the mainland 13. Such an integrated approach has created an ecosystem whose very structure resists easy replication by Western competitors.
The Weaponization of Export Controls
It must be understood that China’s export restrictions are not episodic reactions to market conditions but instruments of strategic leverage, deployed under a dynamic legal framework that permits rapid expansion. In April 2025, Beijing restricted exports of seven rare earth elements and required special licenses for magnets 14; additional controls on five elements followed in October 2025 30. Though some measures have been suspended until November 2026 under diplomatic agreements 17,24,30, the underlying regulatory authority remains intact 30, and the list of controlled commodities can be enlarged on short notice to encompass manganese, tungsten, bismuth, and other materials 23. Critically, these actions are motivated by state policy rather than domestic consumption cycles 27, making the environment inherently unpredictable for downstream purchasers.
The market consequences have been immediate and severe. Prices for magnets outside China now run 1.5 to 3 times those prevailing before the restrictions 16. Dysprosium oxide prices have risen four‑ to five‑fold 16, terbium oxide has trended toward $4,500 per kilogram 16, and yttrium oxide has surged by a factor of roughly 140 16. Supply allocations have been equally stark: Japan received only 4% of its historical dysprosium volumes, while Germany obtained none at all 16, and heavy rare earth exports overall remain about 50% below pre‑control levels 16. Such dislocations exact a direct cost on any enterprise that relies on permanent magnets, lithium‑ion cells, or advanced semiconductor substrates.
The Price of Dependence
For Alphabet Inc., the exposure is systemic even though indirect. The company’s global network of data centers—the engine of Google Cloud and the training ground for large‑language models—depends on rare earth magnets for cooling fans, precision motors, and hard disk drives. Its consumer devices, from Pixel smartphones to Nest thermostats and Chromebooks, incorporate permanent magnets, lithium‑ion batteries, and semiconductor components that rely on gallium and germanium. Hardware under development, including quantum processors, may draw upon helium, niobium, or specialized rare earth compounds 10. In every one of these product lines, China’s bottleneck position in processing translates into heightened cost volatility and physical supply risk.
Export controls on gallium—where a full ban could cost the U.S. economy some $3.4 billion 12—threaten the availability of substrates used in power amplifiers, LEDs, and advanced logic circuits. Restrictions on battery materials, particularly LFP cathode know‑how 5, could impede Alphabet’s adoption of cost‑effective energy storage for data center backup power, even as falling battery costs otherwise stimulate deployment in other sectors 28. Meanwhile, the dynamic character of China’s control list means that materials not currently encumbered could become so with little warning, forcing hardware procurement into a perpetual state of contingency planning.
Pathways to Resilience
The structural gap is not going unnoticed. Western‑aligned projects, while still embryonic, point toward a more diversified future. MP Materials operates the Mountain Pass facility in California 2,3,20, though it still ships ore to China for processing 14, and its Texas magnet plant produces only 1,000 tons annually against China’s 138,000 tons 14. Lynas’s Kalgoorlie plant in Australia is now operational 8, and the Tanbreez project in Greenland offers potential for heavy rare earth extraction 1,8. REalloys is constructing North America’s first integrated heavy rare earth metallization platform 7,8, while firms such as Solvay, Ionic, Ucore, and Cyclic Materials are advancing solvent extraction, recycling, and alternative processing technologies 16,29. Yet sober reflection compels one to acknowledge that the global recycling rate for rare earths remains under 5% 14,29, and China continues to control the lion’s share of oxide separation.
Alphabet can draw a measure of inspiration from Apple, which has achieved 100% recycled rare earth content in its device magnets 8. Emulating that feat across Alphabet’s own hardware lines would require substantial investment in separation and recycling infrastructure, areas where the industry’s capacity is still minimal 14 and costs are elevated. Strategic partnerships with the emerging recycling and processing firms could both buffer supply‑chain risk and advance the company’s sustainability objectives.
Strategic Imperatives
The long‑term implications suggest that Alphabet must view its critical‑mineral dependence as a core strategic challenge, not a peripheral procurement problem. Sustained high costs for magnets and battery materials will pressure margins in the hardware segment already characterized by relatively modest returns. More consequential, any disruption to the build‑out of AI‑optimized data centers—heavily reliant on high‑performance computing components—could slow Alphabet’s competitive momentum at a moment when the AI race brooks no delay 26.
From a strategic perspective, the prudent course involves simultaneous action on multiple fronts: diversifying sourcing, deepening commercial ties with Western‑based processing initiatives, and accelerating the development of internal recycling capabilities. A passive posture, however comfortable in the short term, merely transfers control over the pace and cost of technological development to an external power. The historical record indicates that such dependencies, left unmanaged, ossify into vulnerabilities that economic pressure alone cannot reverse. Alphabet would do well to internalize that lesson before the next wave of export controls renders it academic.
