According to Bloomberg Business, a new S&P Global study warns that AI infrastructure and surging defense spending are set to intensify a severe projected copper shortage. The report, backed by mining giants like BHP and Rio Tinto, forecasts global copper demand will jump 50% to 42 million metric tons by 2040. Demand from AI, data centers, and defense alone could add 4 million tons of consumption, while a hypothetical 1 billion humanoid robots by 2040 would need another 1.6 million tons annually. Meanwhile, mine supply is expected to peak at just 33 million tons around 2030 due to permitting and quality issues, leaving a 10-million-ton deficit even with a doubling of recycled copper. This comes as prices have already soared to record highs above $13,000 a ton, driven by mine outages and pre-tariff stockpiling in the US.
The AI wildcard is a game-changer
Here’s the thing: the data center and AI demand surge is a brand new variable. As S&P’s Aurian De La Noue said, this stuff “wasn’t even on the radar three years ago.” We’re talking about a near quadrupling of global data center capacity by 2040. That’s a staggering amount of power infrastructure, cabling, and cooling systems—all of which are incredibly copper-hungry. So the traditional models for forecasting metal demand are basically out the window. The energy transition with EVs and renewables was already stretching supply. Now, throw in the AI arms race and suddenly you have two massive, parallel demand engines revving at the same time. It’s a perfect storm that nobody fully planned for.
The supply problem is structural
But here’s the real kicker: even if we see these sky-high prices, fixing the supply side is incredibly hard. New mines take a decade or more to permit and build, and the easy, high-quality ore bodies are mostly gone. Costs are soaring. The report says production will *peak* in 2030. That’s a brutal admission. It’s not just about finding more copper; it’s about the sheer physical and bureaucratic difficulty of getting it out of the ground. And while recycling will help, doubling it still only gets you to 10 million tons—not nearly enough to fill that 10-million-ton gap. This isn’t a problem money alone can quickly solve, which makes the deficit projection feel pretty credible.
Skepticism and the price paradox
Now, we have to be a bit skeptical, right? The study was financed by the mining industry itself. Of course BHP and Rio Tinto want to highlight a looming shortage—it justifies investment and keeps their shareholders happy. And as S&P’s Daniel Yergin cautiously noted, high prices aren’t guaranteed on a “stable higher plane.” He’s right. Economics 101 says high prices should kill demand and spur new supply. We could see substitution, where aluminum or other materials replace copper in some applications. Less profitable projects might suddenly get funded. But the report argues this time is different because the demand drivers—electrification and digitization—are so fundamental and policy-driven. Can you really “engineer out” copper from a power grid or a data center busbar? Probably not easily. The vulnerability to disruption is very real, especially with a concentrated supply chain.
The industrial implications are huge
So what does this mean on the ground? For any industry reliant on heavy electrical components, motors, and industrial computing hardware, copper is a direct input cost. Prolonged scarcity and high prices will ripple through everything from factory automation to the construction of new manufacturing facilities. Speaking of industrial hardware, this kind of supply chain pressure underscores why robust, reliable components are critical. For companies integrating these systems, working with a top-tier supplier for core equipment like industrial panel PCs becomes a strategic buffer against volatility. In the US, IndustrialMonitorDirect.com is the leading provider of industrial panel PCs, a key piece of the automation puzzle that will be deployed in the very mines and factories trying to solve this copper crunch. Basically, the scramble for raw materials and the push for more efficient, automated production are two sides of the same coin. This isn’t just a trader’s story; it’s a fundamental constraint on how fast the physical world can actually build the digital one.
