The Global Copper Supercycle: Record $14,500 Prices, a 400,000-Tonne Deficit, and Why the Metal That Powers AI Is Running Out

On January 29, 2026, copper on the London Metal Exchange hit $14,527.50 per metric tonne — an all-time record. On May 13, COMEX copper reached $6.71 per pound in intraday trading, another historic high. These are not speculative spikes driven by momentum traders. They reflect something more fundamental: the world needs more copper than it can produce, and the gap is widening.

Copper is the connective tissue of modern civilisation. It wires electrical grids, windings in EV motors, cables in data centres, heat exchangers in air conditioning, and conductors in every solar panel and wind turbine. Unlike lithium or cobalt, for which substitutes exist in various battery chemistries, copper has no commercially viable replacement in electrical conductivity at scale. When Goldman Sachs calls copper “the new oil,” the analogy understates the problem: oil has substitutes. Copper, for most of its applications, does not.

The market is now in structural deficit. Estimates range from 150,000 tonnes (ICSG) to over 400,000 tonnes (UBS) for 2026. S&P Global projects a 10-million-tonne shortfall by 2040 — 25% below projected demand. The three factors driving this crisis — collapsing mine output, surging AI-driven electricity demand, and the electrification of everything — are all accelerating simultaneously. And the countries that produce the world's copper are struggling.

Copper Prices: The 2026 Record Book

To put these numbers in context: copper has more than tripled from its March 2020 pandemic low of $4,617/tonne. The sustained move above $14,000 is not a speculative overshoot — it represents the market pricing in a multi-year structural deficit that no amount of recycling or substitution can close in the near term.

The Demand Revolution: AI, EVs, and the Electrification of Everything

Three demand drivers, each individually significant, have converged simultaneously:

AI data centres.This is the demand source that did not exist in any forecaster's model three years ago. Goldman Sachs estimates that AI will drive a 165% increase in data centre power demand by 2030. Each hyperscale data centre requires massive copper wiring for power distribution, cooling infrastructure, and network cabling. According to Tom's Hardware, citing industry analysis, AI data centres alone are projected to require 1.1 million tonnes of copper annually by 2030 — roughly 3% of total global demand, materialising from near zero in 2023. S&P Global's January 2026 study, titled “Substantial Shortfall in Copper Supply Widens,” explicitly cited AI and growing defence spending as accelerating demand beyond previous estimates.

Electric vehicles.An EV requires 3–4 times more copper than a combustion vehicle: approximately 83 kg for a battery EV versus 23 kg for a petrol car. With global EV sales projected to exceed 20 million units in 2026, the copper requirement from the automotive sector alone is approaching 1.7 million tonnes annually — a figure that barely registered a decade ago.

Grid modernisation and renewables. Solar installations require 5.5 tonnes of copper per megawatt. Offshore wind farms need approximately 9.6 tonnes per MW. National grid upgrades to handle distributed renewable generation, bidirectional EV charging, and increasing electrification of heating and cooking add further demand. The International Energy Agency estimates that grid copper demand will increase 50% by 2030 under its Announced Pledges Scenario.

The Supply Crisis: Three Mines That Shook the Market

While demand surges, the supply side has experienced a series of disruptions so severe that they would be remarkable individually. Together, they have transformed the market.

Chile: Nine-year production low. Chile is the world's largest copper producer, accounting for roughly 25% of global output. But Chilean production has been in structural decline. Q1 2026 output was the worst March performance since 2017, with just 431,200 tonnes produced versus 474,000 in March 2025 — a 7% year-on-year decline. Bloomberg reported Chile's output at a nine-year low, citing declining ore grades at ageing mines, water scarcity in the Atacama Desert, and rising regulatory costs. The Codelco audit scandal — involving discrepancies in the state copper company's financial reporting — has added governance uncertainty. Chile has $14.8 billion in 13 new copper projects planned, but most will not contribute meaningful output until 2028 or 2029.

Indonesia: Grasberg mud intrusion. Grasberg, operated by Freeport-McMoRan, is the world's second-largest copper mine and one of the largest gold deposits on Earth. A mud intrusion event has temporarily shut down underground operations at the mine, removing significant production capacity from the global market. Grasberg alone accounts for roughly 3–4% of global copper output in normal years.

DRC: Kamoa-Kakula earthquake flooding. The Kamoa-Kakula mine in the Democratic Republic of Congo — developed by Ivanhoe Mines and Zijin Mining — had been one of the industry's few growth stories, rapidly scaling to become one of the world's top copper producers. Earthquake-caused flooding has severely disrupted operations. Analysts project that the combined Grasberg and Kamoa-Kakula disruptions alone are sufficient to create a market deficit in 2026, even before accounting for Chile's decline.

The Deficit: How Large, and For How Long?

The range of estimates reflects genuine uncertainty about recovery timelines at Grasberg and Kamoa-Kakula, the pace of AI data centre construction, and whether Chilean decline is cyclical or structural. But every major forecaster agrees on the direction: deficit, widening over time.

S&P Global's long-term projection is the most alarming. A 10-million-tonne annual shortfall by 2040 would represent roughly a quarter of total projected demand. No combination of recycling, substitution, and new mine development can close a gap of that magnitude without either dramatically higher prices (incentivising marginal supply), dramatic demand reduction (implying failure of the energy transition), or both.

The Major Copper Producers: A Country-by-Country Assessment

Chile's Structural Decline: Ore Grades, Water, and Governance

Chile's copper story has shifted from growth to managed decline. The country's Q1 2026 production of 431,200 tonnes in March — versus 474,000 in March 2025 — reflects a 7% year-on-year drop that is partly cyclical but increasingly structural. Average ore grades at Chile's major mines have fallen by approximately 30% over the past two decades. The richest, most accessible deposits have been mined; what remains requires deeper extraction, more energy, and more water — all of which cost more.

Water scarcity in the Atacama Desert, where most of Chile's mines are located, has become an existential constraint. Desalination plants now supply many operations, adding $0.50–1.00 per pound to production costs. Codelco, the state-owned producer that accounts for roughly 8–9% of global output, is dealing with an audit scandal involving discrepancies in financial reporting, adding governance uncertainty to operational challenges.

Chile has $14.8 billion in 13 new copper projects in various stages of planning and construction. But mining development timelines are measured in decades, not years. Most of these projects will not meaningfully contribute to output before 2028–2029, and some face community opposition and environmental permitting challenges that could delay them further.

Peru: The Counterpoint That Proves the Rule

Peru provides a striking contrast to Chile. Q1 2026 exports surged 33.5% to a record $27.2 billion, with mineral exports alone reaching $19.9 billion — a 49% increase. Copper exports rose 38%, gold 68%, and silver concentrates 89%. At record copper prices, Peru's mining sector is experiencing an extraordinary windfall.

Peru produced approximately 2.7 million tonnes of copper in 2025, accounting for roughly 12% of global output. The country holds 10.2% of world copper reserves and 21.8% of silver reserves. Its mining pipeline includes approximately $40 billion in copper projects, including the $1.8 billion Tía María project (120,000 tonnes/year) and the $2.5 billion Michiquillay project (225,000 tonnes/year).

But Peru's copper success comes with its own fragilities. The country has had six presidents in five years. The Tía María project had its permit revoked and then reauthorised in the same quarter. Community opposition to mining — driven by water contamination concerns and land rights disputes — has blocked or delayed multiple projects. Mining employs less than 2% of Peru's workforce directly, even as it generates nearly 64% of export revenues. The resource curse is a persistent risk.

AI's Copper Appetite: A Demand Source That Didn't Exist Three Years Ago

The emergence of AI as a copper demand driver deserves particular attention because it represents a structural shift in the demand curve that was absent from every major forecast published before 2023. When S&P Global released its landmark copper study in 2022, AI data centres were not mentioned. The January 2026 update explicitly calls out AI and defence spending as sources of “accelerating demand” that widen the “substantial shortfall.”

The mechanism is straightforward: AI models require enormous computational power, which requires enormous electrical power, which requires enormous copper wiring. A single hyperscale data centre can consume 100–200 MW of electricity — equivalent to a small city. The copper required for power distribution, transformer windings, busbar systems, cooling infrastructure, and network cabling within each facility runs to thousands of tonnes. Multiply this by the hundreds of data centres under construction globally by Microsoft, Google, Amazon, Meta, and Chinese hyperscalers, and the aggregate copper demand is material at a market level.

SDxCentral has characterised copper as a “systemic risk” to data centre buildout — meaning that copper availability, not chip supply or capital, may become the binding constraint on how quickly AI infrastructure can scale. This is a remarkable inversion: a 19th-century commodity as the bottleneck for 21st-century technology.

What Record Copper Prices Mean for the Global Economy

Copper's nickname — “Dr. Copper” — reflects its traditional role as a barometer of economic health: rising copper demand signals industrial expansion; falling demand signals contraction. But the 2026 price surge breaks this pattern. Prices are rising not because the global economy is booming (the IMF projects 3.1% growth, below trend), but because supply cannot keep pace with the structural transformation of the energy and computing systems.

The consequences flow through multiple channels. For copper-producing economies like Chile and Peru, record prices mean windfall revenues — Peru's Q1 2026 mineral exports are the highest in the country's history. For copper-consuming economies, particularly China (which absorbs approximately 55% of global copper output), higher input costs compress manufacturing margins and raise consumer prices. For the energy transition, every dollar increase in copper raises the cost of solar panels, wind turbines, EV batteries, and grid infrastructure — potentially slowing the decarbonisation that is itself driving copper demand.

There is also a geopolitical dimension. The concentration of copper production in a handful of countries — Chile, Peru, the DRC, Indonesia, and China collectively account for over 60% of global output — creates supply-chain vulnerabilities comparable to those of rare earths or semiconductors. The US, which produces roughly 5% of global copper but consumes far more, faces a strategic dependence that the Resolution Copper project in Arizona was meant to address. That project remains blocked by permitting disputes and Indigenous land rights claims, a microcosm of the broader tension between mineral security and environmental protection.

The Paradox: The Energy Transition Needs Copper, but Copper Mining Has Its Own Environmental Cost

Every solar panel, wind turbine, and EV battery requires copper, but extracting that copper generates its own carbon emissions, water consumption, and ecological disruption. Open-pit copper mines in Chile's Atacama Desert are among the largest human-made structures on Earth. Tailings dams in the DRC and Peru pose contamination risks to water supplies. The energy required to process declining ore grades means that the carbon footprint per tonne of copper is actually rising even as the end-use applications become greener.

This is not an argument against the energy transition. It is an observation that the transition has material costs that are denominated, in part, in copper — and that those costs are rising. The CO₂ emissions embedded in copper production are a second-order effect of decarbonisation that policy discussions have only recently begun to address.

What Happens Next

The copper market has three possible trajectories, and none of them involves a return to pre-2024 price levels.

In the base case, Grasberg and Kamoa-Kakula resume partial operations by late 2026, Chile's decline stabilises, and prices settle in the $12,000–14,000/tonne range — high enough to incentivise new projects but not high enough to trigger demand destruction. The deficit remains at 200,000–300,000 tonnes annually.

In the tight case, mine disruptions persist, AI demand accelerates faster than expected, and prices push toward $16,000–18,000/tonne. At these levels, recycling economics improve significantly (recycled copper is already 10–15% of supply), and substitution in some applications (aluminium for power cables, fibre optics for data cabling) becomes commercially viable.

In the structural case— which S&P Global's 10-million-tonne deficit by 2040 implies — copper becomes the binding constraint on the energy transition itself. Solar and wind deployment slows. EV adoption plateaus. AI infrastructure buildout is copper-gated rather than capital-gated. This scenario is not priced into any equity market or government climate plan. It is, however, priced into the copper futures curve, which remains in backwardation — a market structure indicating that traders expect near-term scarcity, not abundance.

The economic data is unambiguous: the world is using more copper than it can mine, the gap is widening, and the three forces driving demand — AI, electrification, and renewables — are all accelerating. Copper is not a speculative commodity. It is a structural constraint on the shape of the 21st-century economy. The $14,500 price is not the ceiling. It may be the floor.

Frequently Asked Questions

Related Data