The WTO Warns The Iran Conflict Could Halt AI Progress

The most consequential threat to AI investment right now isn't regulation, competition, or an underperforming model. It's the price of electricity.

The World Trade Organization's latest Global Trade Outlook identifies the war in Iran and its effects on energy and fertilizer costs as the primary risk to the global economy in 2026. But the report raises a more specific concern for the technology sector: sustained high oil prices could, in the words of WTO chief economist Robert Staiger, "put a crimp on the AI boom."

The mechanism is direct. AI infrastructure is extraordinarily energy-intensive. Data centers running large model training and inference at scale consume power at a rate that makes energy costs a material line item for every major AI company. When that cost rises and stays elevated, the economics of AI investment change.

The Scale of What's at Stake

The WTO's framing of AI investment risk lands differently when you see the numbers behind it. In the first three quarters of 2025, approximately 70% of all investment growth in North America was attributable to AI-related goods. For context, in the three years preceding the 2008 housing crash, real estate accounted for 30% of US investment growth.

That comparison is not made casually. The WTO is signaling that AI investment has become the dominant engine of North American capital allocation — and that the concentration of that investment in a small number of very large firms, combined with technology that remains, as Staiger put it, "ultimately unproven in terms of how much it can deliver," creates meaningful uncertainty about its trajectory.

The AI boom helped absorb a significant portion of the damage to global trade in 2025 caused by Trump's tariffs, which pushed US tariff rates to their highest level in decades. Despite that pressure, world goods trade expanded 4.6% last year, with Asian export performance carrying much of the weight. The WTO now expects the growth rate to fall to 1.9% in 2026 — and a sustained energy price shock would knock an additional 0.5% off that figure, while also threatening food security through higher fertilizer costs.

Energy as AI Infrastructure Risk

The relationship between energy costs and AI viability is underappreciated in most business planning conversations. The prevailing narrative around AI investment focuses on model capability, competitive differentiation, and return on deployment. Energy is treated as an operational cost — present, but manageable.

The WTO report highlights that energy is not just an operational variable. It is a structural dependency for the entire AI build-out. The hyperscaler data centers, the GPU clusters, the inference infrastructure running behind every AI product in market — all of it draws power at scale. When the cost of that power is driven by geopolitical events outside any company's control, it introduces a risk vector that doesn't appear in most AI business cases.

The firms most exposed are those running the largest training workloads and the most inference-heavy deployments — which are, not coincidentally, the same firms that account for the bulk of that 70% investment figure. Microsoft, Google, Amazon, Meta, and the infrastructure providers serving them are all operating in an environment where energy cost projections made six months ago may no longer hold.

What This Means for the Businesses Downstream

For organizations that are consumers of AI rather than builders of it — which describes most companies deploying AI in marketing, operations, and customer experience — the energy risk doesn't hit directly. It travels through pricing, availability, and the pace at which AI capabilities are developed and released.

If sustained energy costs slow the rate of infrastructure build-out, the capacity constraints that already affect access to frontier models could tighten. If they pressure the financial returns on AI investment, the timeline for large AI companies to demonstrate commercial viability shortens — and the decisions they make about which products and capabilities to prioritize may shift accordingly.

None of this unfolds overnight. But the WTO's warning is a useful reminder that AI development does not exist outside the physical economy. It runs on chips that require rare materials, in buildings that require enormous amounts of power, in a geopolitical environment that can reprice both without notice.

The organizations building AI strategy with durable assumptions baked in — rather than optimistic projections about stable infrastructure costs — will be better positioned when those assumptions are tested.

If your organization is thinking through AI strategy and its real-world cost structures, Winsome Marketing's growth team can help stress-test the plan.