The AI industry has been modeled for many risks. A geopolitical war shutting down 20% of the world's oil supply was not high on most lists.
The World Trade Organization issued a stark warning this week: if energy prices from the US-Iran conflict remain elevated for the rest of 2026, global GDP growth would fall from a baseline of 2.8% to 2.5% - and the AI investment boom that has been propping up much of the global economy could be among the casualties. WTO chief economist Robert Staiger told the Guardian plainly: "If the price of energy continues to be elevated for the whole year, that could put a crimp on the AI boom."
The Mechanism Is Straightforward
Iran's blockade of the Strait of Hormuz - the narrow passage through which roughly 20% of the world's seaborne oil and LNG travels daily - has driven Brent crude to around $90 a barrel, with some analysts warning of $120 or higher if the disruption persists. Gas prices in the US hit $4 per gallon on March 31. European natural gas benchmarks surged 30% in a single day after Israeli strikes hit Iran's South Pars gas field and Iran retaliated against Qatar's Ras Laffan LNG facility - the largest in the world.
AI data centers run primarily on natural gas. They are among the most energy-intensive structures humanity has ever built - a single AI server rack requires 40 to 100 kilowatts of power compared to 5 to 15 for a traditional rack. US data centers already consume roughly 4.4% of national electricity, with the IEA projecting that half of all US electricity demand growth over the next five years will come from data centers alone.
The Financing Problem
Staiger's concern goes beyond operating costs. The AI buildout is financed by enormous amounts of debt - OpenAI's $122 billion round closed the same week Iran blocked the strait. Higher oil prices feed inflation, which delays interest rate cuts, which increases the cost of capital for infrastructure spending that is already stretched thin. Goldman Sachs estimates data center electricity demand will add 0.1 to 0.2 percentage points to core US inflation annually. Private credit firms that have financed much of the data center construction are already under pressure.
The WTO's own data shows AI-related goods - semiconductors, servers, and telecom equipment - rose 20% year over year in the first half of 2025 and accounted for nearly half of all merchandise trade expansion in that period. That supply chain runs through ships that transit the Gulf, uses specialty gases sourced from the region, and depends on stable financing conditions that energy shocks erode.
The Uncomfortable Strategic Reality
From my four years watching executives implement AI across their organizations, the hard truth here is that the AI buildout was conceived as a peacetime project. The scaling laws that drove it assume abundant energy, expanding chip supply, and stable financing. All three are now under pressure simultaneously. Companies will not abandon AI investment - the strategic imperative is too strong. But timelines will slip, cost projections will widen, and the economics that looked inevitable six months ago look considerably less certain today.
