As of March 2026, the AI revolution faces significant supply chain constraints. Key resources such as copper, electricity, and memory chips, essential for AI operations, are becoming bottlenecks. To sustain AI’s current growth trajectory, the world would need to mine the same amount of copper over the next 18 years as it has in the last 10,000 years combined. Current copper investments have more than doubled, while major AI companies like Amazon, Meta, and Microsoft have seen only 1% gains.
Demand for energy to power data centers is straining local grids, with electricity prices surging by up to 267% over five years in some areas. Similarly, the shortage of memory chips (DRAM) could persist for years, as nearly 100 gigawatts of new data centers are expected to come online, but only 15 gigawatts of DRAM supply will be available in the same timeframe. This situation raises significant concerns about the future of AI growth.
Key companies tied to the physical infrastructure of AI, such as those involved in copper mining, energy generation, and memory manufacturing, are well-positioned to benefit in this tightening market. Their ability to control these crucial chokepoints could determine their profitability as the AI boom unfolds.
