The International Energy Agency’s Global Critical Minerals Outlook 2026 reduced its projected copper supply gap for 2035, lowering the estimate from about 30% in last year’s report to roughly 25% this year. After a decade in which the projected gap had only widened, the change marks measurable progress from projects that are advancing.

For copper demand tied to electrification, data centers and modern infrastructure, each percentage point represents millions of tonnes of metal and billions of dollars in investment. That makes the direction of the IEA revision important, even if the remaining shortfall is still large enough to keep pressure on the full mine-to-metal chain.

Most of the improvement comes from existing mines

The narrower gap is driven almost entirely by expansions and life extensions at mines that are already operating. These include the Kisanfu and Lumwana expansions in Africa, the Highland Valley life extension in Canada and the Antamina extension in Peru.

Those projects are important, but closing the gap through expansions and life extensions is comparable to drawing down savings rather than increasing income. Savings are finite, and genuinely new supply that would replenish the account is not arriving at the required pace.

Resolution Copper in Arizona, one of the largest undeveloped copper deposits in North America, remains so tied up in litigation that the IEA excluded it entirely from both its base case and high case forecasts. The supply gap can close only if the obstacles blocking such deposits are resolved, not by forcing them into production.

The improvement in the IEA report is real. It also shows an industry relying heavily on assets it already has.

A critical processing step is under strain

The fragility lies in a step that is not widely visible outside the mining industry.

Before copper becomes metal, ore is concentrated by flotation and then smelted. Most sulfur is removed during concentration, while the remainder is driven off in the furnace. Smelting is a choke point: nearly every tonne of the world’s primary copper passes through a smelter, and that stage is now under both financial and physical pressure.

On the financial side, the benchmark fee to process a tonne of copper concentrate settled at zero dollars in early 2026, the lowest level ever agreed, according to the IEA. Spot fees have been negative since 2024. These fees are the price miners pay smelters to turn concentrate into metal, so a zero or negative number signals that smelters are competing intensely for too little concentrate. Since 2005, China has built more than 90% of the world’s new copper smelting capacity, increasing its share from about 15% to roughly half of global capacity. That expansion has outpaced available copper concentrate so significantly that processing fees have been competed down to nothing.

Smelters outside China now operate at below 70% of capacity, while Chinese smelters run near 85%. Many facilities remain open only because they can sell byproducts such as gold, silver and sulfuric acid that accompany the copper. If prices for those byproducts fall, smelters, particularly outside China, face a higher risk of closure.

The physical strain is also becoming visible as furnaces go offline. Some outages reflect normal aging, especially among the many smelters outside China that are more than 40 years old and have equipment wearing out. But when processing fees are zero and there is little indication they will recover soon, a smelter has limited reason to keep operating at a loss.

Bringing maintenance forward and taking a furnace down for an extended overhaul can become a rational way to wait out a cycle with no upside. While the furnace is down, however, the copper has nowhere to go.

Zambia illustrates the cost

That problem appeared this year in Zambia, one of the countries the IEA credits with helping to narrow the supply gap.

Zambia imposes a 10% duty on exports of raw copper concentrate. The policy reflects an ambition shared by many copper-producing countries: process concentrate domestically and capture the value of finished metal instead of exporting it as rock.

In June, for the second time since August 2025, Zambia suspended that duty on more than 270,000 tonnes of concentrate because its own smelters were down for extended maintenance and could not process the copper, as Reuters and Mining.com reported. A country that had written its processing ambition into law had to export unfinished copper and hand the value to another smelter because the step needed to capture that value had failed.

That single event captures the broader issue. Mine supply is not the same as usable metal supply if the processing system cannot absorb the concentrate. When the smelting link gives way, the value of copper leaves on a ship.

Protecting the progress

If the weak link is a centralized and financially strained smelter, the solution is not simply a larger smelter somewhere else. The alternative is a processing route that a producing country can operate at its own mine without a furnace in the chain.

Recovering copper through new leaching methods, ideally while converting sulfur into a saleable product as smelters do, could bring projects to market more quickly. Such an approach would avoid the need to build, staff or shut down a smelter for repairs. A small number of companies, including Still Bright, are working to prove that route at commercial scale.

None of this reduces the significance of the progress recorded in the IEA report. But supply built on drawn-down savings and routed through a processing step that can go dark when the economics turn is not supply the world can fully count on, nor a future its owners can confidently build around. Protecting the good news means reducing dependence on the step that continues to break.

Ranulfo (Randy) Allen, PhD, MBA, is the co-founder and CEO of Still Bright, a seed-stage deep-tech company using proprietary electrochemical reductive leaching of rougher concentrates to recover copper.