Silicon Valley gets most of the headlines, but the AI boom is quietly being wired together with a metal that has been in the ground for millions of years. Copper, the same red metal that lines your home's electrical panel and runs through your car's engine, has become one of the most structurally important commodities of the current technology cycle, and the market is only beginning to price that in.
Prices for COMEX copper have risen from around $4.10 per pound a year ago to approximately $5.65 per pound as of early 2026, approaching all-time highs on both the COMEX and the London Metal Exchange. But the forces behind that rally are not speculative. They are structural, they are accelerating, and for US investors who have not positioned for it yet, the window to get ahead of the trade may be narrowing.
The numbers are staggering once you start adding them up. Large AI campuses are now routinely designed around power blocks of 50 to 150 megawatts, and industry estimates place copper consumption at roughly 27 to 33 tonnes per megawatt of installed capacity. By that math, a single 100-megawatt site can absorb several thousand tonnes of copper before accounting for the upstream grid reinforcements needed to supply it. BHP's own case studies cite more than 2,000 tonnes for an 80-megawatt-class deployment.
Multiply that across the hyperscale ambitions of Microsoft, Google, Amazon, and Meta, all of which are spending hundreds of billions on infrastructure buildouts, and you get a demand signal that the copper market has simply never seen before. Goldman Sachs estimated that AI will drive a 165% increase in data center power demand by 2030, and all of that power needs copper to move it, from the grid to the transformer to the rack.
According to S&P Global's landmark study on copper in the age of AI, data centers are driving massive investments in both direct copper use and in the electric grid infrastructure that supports them. The report puts AI and data center spending at the center of what S&P Global Vice Chair Daniel Yergin described as a convergence of demand that supply simply is not on track to meet: grid expansion, renewable generation, AI computation, electric vehicles, and defense, all scaling simultaneously.
Copper's supply problem did not appear overnight. Ore grades at legacy mines have fallen by roughly 40% since 1991, leaving operators to either squeeze diminishing returns out of aging sites or navigate years-long permitting battles to access newer deposits. The International Energy Agency has flagged that existing and planned mines can meet only about 70% of projected 2035 demand, and Wood Mackenzie has estimated that a refined-copper deficit of 304,000 tonnes already materialized in 2025, with a wider gap expected in 2026.
The market is also absorbing a pair of unplanned production shocks. Disruptions at the Grasberg mine in Indonesia and the Kamoa-Kakula mine in the Democratic Republic of Congo have meaningfully reduced global refined copper output at precisely the moment demand is accelerating. BloombergNEF expects the cumulative copper locked into data centers to surpass 4.3 million tonnes by 2035, and AI-powered facilities are projected to average around 400,000 tonnes of copper demand annually over the next decade, peaking near 572,000 tonnes in 2028.
S&P Global's research projects a potential supply deficit that could reach 10 million metric tons by 2040, with global demand set to surge by 50% relative to current levels. A study of this scope is not a niche commodity report; it is a macro-level warning signal for markets that depend on electrification, and for investors, it raises an obvious question about how to position.
Trade policy has added a layer of complexity to the copper narrative that US traders should understand before entering a position. In August 2025, the US imposed a 50% tariff on copper imports, a move designed to shore up domestic production. The tariff announcement triggered a sharp COMEX-to-LME price spread, which briefly widened to nearly 30% in the summer of 2025, as traders stockpiled US-inventoried copper ahead of the restriction. Copper prices fell sharply in the immediate aftermath, with some measures showing a decline exceeding 20% before the market stabilized.
By early 2026, that spread has narrowed significantly, and both COMEX and LME prices are now rising in tandem. Analysts view this convergence as a healthier signal than the 2025 divergence because it suggests the current price strength reflects genuine underlying demand rather than inventory positioning ahead of policy. China's trade behavior has added another wrinkle, with Chinese refined copper exports surging sharply in late 2025, complicating the global supply-demand calculus.
The tariff policy does benefit domestic miners such as Freeport-McMoRan, but it can also introduce short-term pricing distortions that create volatility. Investors should treat tariff developments as an ongoing variable in any copper position rather than a fixed tailwind.
The structural setup for copper is clear enough. What differs is how individual investors choose to express that view given varying risk tolerances and time horizons.
Freeport-McMoran (NYSE:FCX) remains the most direct equity play on copper for US investors. The company reported strong Q1 2026 earnings of $881 million in net income and $6.23 billion in revenue even as it navigates ongoing operational challenges at its Grasberg mine in Indonesia. Wall Street analysts carry price targets ranging from $66 to $75 per share and the company's management has guided for average annual copper production of approximately 1.6 billion pounds between 2027 and 2029 as Grasberg works through its recovery plan.
Freeport's leverage to copper prices is well-documented. When copper moves higher, Freeport's margins expand quickly because its production costs are largely fixed. The risk, equally well-documented, is that the Grasberg mine continues to run below full capacity and that geopolitical friction in Indonesia adds uncertainty to its long-term operating rights. The company has signed an agreement to extend those rights but the process has required transferring additional government stakes in its Indonesian subsidiary.
For investors who want diversified exposure across the copper mining sector rather than a single-stock bet, Global X Copper Miners ETF (NYSE:COPX) is the most widely followed option. The fund holds 41 copper mining stocks with its top five positions including Glencore (5.3%), Teck Resources (5.2%), BHP Group (5.1%), Antofagasta (4.9%), and Zijin Mining (4.9%). Its expense ratio is 0.65%, and as of early May 2026, the ETF has gained roughly 116% over the trailing twelve months—a performance that reflects just how aggressively institutional money has been flowing into the copper space.
One meaningful caveat with COPX—and with copper mining ETFs generally—is that most of these companies produce other metals alongside copper including gold molybdenum and iron ore. That means the fund's performance will not track copper prices in a pure 1-to-1 fashion. In periods where copper outperforms the broader metals complex miners can either lag or lead depending on their cost structures and the mix of production. Investors should size their positions with that in mind.
iShares Copper and Metals Mining ETF (NASDAQ:ICOP) offers a slightly different composition holding around 45 mining companies with top positions in Anglo American (9%), BHP Group (8.5%), Grupo México (7.8%), and Freeport-McMoRan (7.7%). Its expense ratio of 0.47% is lower than COPX's which matters compounded over a multi-year hold period. For investors who prefer a slightly more defensive lower-cost entry into the copper mining theme ICOP is worth comparing directly against COPX before committing capital.
Copper's bull case is compelling enough on data centers alone but AI infrastructure is only one of the demand vectors pulling at a constrained market. Electric vehicles require roughly 2.9 to 4 times more copper than conventional gasoline-powered cars; global EV sales continue climbing. Renewable energy infrastructure—including wind turbines, solar installations, and grid upgrades needed to route power from generation to consumption—is another massive source of demand growing in parallel with AI buildout.
BloombergNEF projects copper consumption from power transmission and wind energy to nearly double by 2035. When you layer that demand on top of data center and EV curves, supply math becomes increasingly difficult. Analysts at Lean Research have flagged a potential price range of $11,000 to $12,500 per metric ton for 2026 as a base case; some estimates extending toward a 10-million-tonne cumulative deficit by 2040.
Copper's role in all of this is non-discretionary. Hyperscalers cannot build AI campuses without it. Grid operators cannot expand transmission capacity without it. Electric vehicles cannot be manufactured without it. In a world where every major growth industry is converging on the same material at the same time, the supply-demand arithmetic points in one direction.
There are a few variables that could interrupt or delay the copper trade, and investors would be well-served to monitor them. Chinese demand remains the most important swing factor in the global copper market. A sustained slowdown in Chinese industrial activity or a collapse in its domestic EV market could meaningfully soften the demand outlook in the near term even as structural AI and electrification tailwinds remain intact over longer horizons.
US tariff policy is also worth tracking closely. The 50% import tariff already in place has skewed domestic inventory dynamics and any further escalation or conversely any reversal in trade posture toward copper-producing nations could trigger sharp price movements. Investors in domestic-focused names like Freeport benefit from the tariff regime but pure commodity futures exposure would be more vulnerable to a policy pivot.
Finally, the pace at which new mine development comes online deserves attention. Several large copper projects are in various stages of permitting and development; faster-than-expected ramp-ups could narrow projected deficits. The Arizona-based Resolution Copper project has been cited as one potential domestic supply source though its path through regulatory approval remains lengthy.
Copper is not a crowded trade yet but catalysts are becoming harder to ignore. The metal powering the AI revolution is facing a structural mismatch between the pace of demand growth and the pace at which mining industry can respond. For US investors,the question is not really whether copper matters to next decade of growth,but whether they want positioned before market finishes figuring that out.