Copper's long-term demand narrative is being reshaped by artificial intelligence infrastructure. Major cloud providers have announced combined capital expenditures exceeding $100 billion annually for data center expansion through 2026, with each facility requiring 3,000-8,000 tonnes of copper for power distribution, grounding, and cooling systems. Goldman Sachs Research estimates AI-related copper demand will reach 1.2 million tonnes by 2030, up from approximately 400,000 tonnes in 2025, representing a 200% increase over five years.
The structural demand story is reinforced by power grid investment. Global grid infrastructure spending is projected to exceed $600 billion annually by 2030, driven by renewable energy integration and grid modernization. Copper intensity per gigawatt of grid-connected renewable capacity is approximately 3-5 times higher than conventional generation, creating a durable demand floor that persists across economic cycles. Defense spending on electrification and munitions adds another 300,000-400,000 tonnes of annual demand by 2028.
Goldman Sachs Research forecasts the LME copper price in a range of $10,000-$11,000 per tonne for the remainder of 2026, as strong demand growth from grid and power infrastructure offsets tariff uncertainty. The bank's long-term forecast calls for $15,000 per tonne by 2035, above consensus expectations. JP Morgan projects a 2026 average of $12,075 per tonne, with potential upside to $12,500/t in Q2. Bank of America has raised its 2026 forecast to approximately $11,313 per tonne, citing mine disruptions and historically low exchange inventories.
Long-term contracts should incorporate price escalation clauses tied to LME, as the structural demand outlook supports elevated pricing. Evaluate fixed-price agreements for H2 2026 while forecasts remain in the $10,000-12,000/t range. The AI-driven demand inflection starting in 2027 will create sustained upward pressure.