Copper Shortage Emerges as Systemic Risk to Global Data Center Expansion, S&P Global Warns
January 9, 2026 The breakneck pace of artificial intelligence infrastructure buildout is colliding with a fundamental physical constraint: a looming global shortage of copper. A new report from S&P Global warns that this critical metal's supply deficit poses a "systemic risk" not only to data center construction but to broader technological advancement and economic growth, adding to existing strains on memory chips and specialized components. According to the study, a projected supply deficit could reach 10 million metric tons by 2040, with overall demand set to surge by 50 percent compared to current levels. The data center industry is a significant and growing contributor to this strain. Copper demand from data centers is forecast to increase from 1.1 million metric tons in 2025 to between 1.7 and 2.7 million metric tons by 2040, with AI training facilities alone expected to account for 58 percent of total data center copper demand by 2030. The report suggests that combined demand from AI data centers and defense spending could represent an additional four million metric tons. "Copper is the great enabler of electrification, but the accelerating pace of electrification is an increasing challenge for copper," said Daniel Yergin, vice chair of S&P Global, who co-chaired the study. He noted that economic demand, grid expansion, renewable generation, AI computation, and digital industries are "scaling all at once, and supply is not on track to keep pace." The metal's role in digital infrastructure is pervasive, found in power delivery and cooling systems, Ethernet cables, transceiver modules, and chip-level circuitry above the 10-nanometer node. While a shift to fiber optics for longer interconnects may reduce copper intensity by four to five metric tons per megawatt, this is marginal compared to the overall 30-40 metric tons per megawatt required for a typical non-cryptocurrency data center. Nvidia CEO Jensen Huang has previously stated that "we should stay with copper for as long as we can," acknowledging its entrenched role despite ongoing research into photonics alternatives. This copper crunch compounds existing supply chain vulnerabilities. The industry is already grappling with a severe shortage of memory chips, driven by massive procurement deals from AI firms like OpenAI, which has led to panic buying and price hikes. Furthermore, Nvidia's aggressive procurement of electro-absorption modulated lasers (EMLs) for its advanced networking solutions risks creating another bottleneck, with wait times for related components potentially extending beyond 2027. Aurian De La Noue, executive director for critical minerals and energy transition consulting at S&P Global Energy, emphasized the need for strategic planning: "Multilateral cooperation and regional diversification will be crucial to ensure a more resilient global copper system – one commensurate with copper's role as the linchpin of electrification, digitalization, and security in the age of AI." The findings underscore that the race to build AI-capable infrastructure is increasingly a race for the raw materials that make it physically possible. Source: datacenterdynamics
