Global trade is having a surprisingly good year. The reasons may give supply-chain executives more cause for concern than celebration.
On Thursday, the World Trade Organization raised its forecast for merchandise trade growth in 2026 from 1.9% to 3.9%, an extraordinary revision for a year marked by trade barriers, military conflict and disruptions to some of the world’s most important shipping routes.
The explanation lies largely in the warehouses, semiconductor factories and data centers supporting the artificial intelligence boom.
According to the WTO, products used to build AI infrastructure accounted for 47% of global merchandise trade growth in value terms during the first half of 2026. Trade in these products increased 67% from a year earlier.
The numbers suggest that globalization is proving remarkably resilient. They also raise an uncomfortable question: how healthy is global trade when so much of its growth depends on one investment boom?
The AI Boom Has Become a Trade Boom
Artificial intelligence is often discussed as a software revolution, but its immediate economic effects are decidedly physical.
Every new data center requires semiconductors, servers, networking equipment, cooling systems and electrical infrastructure. Much of that equipment crosses international borders several times before reaching its final destination, drawing on specialized manufacturing ecosystems concentrated in East and Southeast Asia.
The scale of the resulting trade flows has surprised even economists.
Research published by DHL and New York University’s Stern School of Business on October 7 found that AI-enabling products accounted for 42% of global goods trade growth in 2025. Their contribution reached 76% during the first quarter of 2026, according to WTO and OECD analysis.
DHL Express Chief Executive John Pearson summarized the shift: “The biggest story in global trade right now is AI – not tariffs.”
That assessment challenges the prevailing narrative of a world retreating from globalization.
Companies may be moving production, governments may be raising tariffs and geopolitical tensions may be escalating. Yet international trade continues expanding because the technology industry’s demand for physical infrastructure is growing faster than many of those pressures can restrain it.
The Numbers Conceal a Two-Speed Economy
The headline growth figures become less reassuring when examined by sector and geography.
The WTO expects Asian merchandise exports to increase 9.9% in 2026. European exports, by contrast, are forecast to decline slightly, while Middle Eastern exports are expected to contract 17.2%.
The disparity reflects more than differences in economic performance. It shows how unevenly countries participate in the industries currently driving international trade.
Asia dominates the production of AI-enabling hardware, while North America provides much of the demand for new computing infrastructure. Countries outside those networks may see relatively little benefit from the boom.
There is also an important distinction between trade value and physical volumes.
The WTO reported that global merchandise trade increased 15% in dollar terms during the first half of 2026, while actual volumes rose only 3.5%. Higher energy prices and the growing importance of expensive technology products contributed to the difference.
For logistics companies, this matters considerably. A shipment of advanced semiconductors can represent enormous trade value without generating anything approaching the physical freight demand of steel, agricultural commodities or consumer goods.
Global trade can therefore become more valuable without producing an equivalent increase in transportation demand across every sector.
Supply Chains Are Adapting Faster Than Expected
The WTO report also reveals another important development.
Global supply chains have proved considerably better at adapting to disruption than many observers anticipated.
Middle Eastern crude oil exports declined approximately 24% during the first half of the year, while liquefied natural gas exports fell 47%. Yet additional shipments from other producing regions limited the contraction in global exports to approximately 6% for crude oil and just 1% for LNG.
Fertilizer supply chains demonstrated similar adaptability as alternative producers helped compensate for severe regional disruptions.
These adjustments are evidence of substantial operational resilience.
But they come at a cost.
Changing suppliers, extending transport routes and securing alternative capacity consumes money and resources. The WTO’s assessment of services trade reflects some of those pressures: it reduced its 2026 growth forecast from 4.8% to 3.3%, partly because higher fuel costs and disrupted transportation networks have damaged travel and transport services.
Goods trade is outperforming expectations while parts of the infrastructure moving those goods remain under considerable strain.
Globalization Is Surviving but Changing Shape
A second finding from DHL’s new research complicates the picture further.
Direct trade between the United States and China has fallen from 3.5% of world trade at its 2015 peak to just 1.6% during the first five months of 2026.
That might appear to confirm that the world’s two largest economies are successfully reducing their dependence on one another.
Yet the study finds that products imported into the United States through other countries increasingly contain Chinese components and materials. When those indirect relationships are considered, American dependence on Chinese production has declined only slightly through 2024, the latest year with available data.
The implication is important.
Supply chains are changing their routes faster than they are changing their underlying industrial dependencies.
A semiconductor assembled in one country may depend on equipment, materials and intermediate components manufactured elsewhere. The final shipping origin tells only part of the story.
As technology supply chains become more specialized, understanding these upstream dependencies becomes increasingly important.
The Next Shock Could Come From Falling Demand
The greatest uncertainty may not be another tariff, port closure or shipping disruption.
It may be a change in the investment assumptions underpinning the AI boom.
Spending on data centers and computing infrastructure is concentrated among a relatively small number of technology companies. Their procurement decisions influence semiconductor production, equipment manufacturing, logistics capacity and investment plans throughout the supply chain.
If those companies slow their expansion, the consequences could spread rapidly across the industrial ecosystem supporting them.
The WTO explicitly identifies weaker AI investment as a risk to its trade outlook.
This does not mean an AI investment downturn is imminent. The current expansion could continue for years, creating substantial opportunities for manufacturers, carriers and logistics providers.
But it does mean that an increasingly important part of global trade is exposed to decisions being made by a comparatively small group of companies.
That is a different kind of concentration risk from dependence on a particular country or shipping corridor.
Global Trade Has Found a New Engine
The WTO’s latest forecast offers compelling evidence that globalization remains economically powerful.
Companies continue to source internationally, supply chains are adapting to geopolitical disruption and technology investment is creating substantial new trade flows.
Yet the growth figures also reveal a structural imbalance.
The global trading system is becoming increasingly dependent on the continued expansion of a technology sector whose long-term investment requirements remain uncertain.
For supply-chain executives, the distinction matters. Rising global trade volumes do not necessarily indicate broad economic strength, just as record investment in one industry does not guarantee stable demand elsewhere.
The world has found a powerful new engine for international commerce.
It remains to be seen how smoothly the rest of the economy will run if that engine slows.
Practitioner Takeaway
Companies should distinguish between trade growth driven by broad industrial demand and growth linked to the AI investment cycle.
This is particularly important when assessing logistics capacity, supplier investment and long-term demand forecasts. A market experiencing strong aggregate growth may still contain considerable weakness outside AI-related industries, while suppliers several tiers removed from technology manufacturers may be more exposed to a slowdown in data-center investment than their immediate customer relationships suggest.
