The world economy has absorbed the latest energy and geopolitical shock without falling into the severe downturn many feared. But the resilience is uneven—and increasingly dependent on a narrow engine: investment linked to artificial intelligence and technology manufacturing.
The OECD’s latest Economic Outlook projects global growth of 2.9% in 2026 and 3.0% in 2027, down from 3.4% in 2025. The organization says sizeable oil inventories, additional supply from outside Gulf economies and government support helped cushion the effects of the conflict in the Middle East. Strong AI-related activity also supported investment, production and trade.
Why the shock has not become a global recession
The immediate transmission mechanism from conflict to the wider economy was energy. A sustained interruption to Gulf supplies could have raised transport, manufacturing and household costs simultaneously. Instead, inventories and alternative production provided a buffer. Governments also limited some of the damage through temporary fiscal measures.
That buffer matters because energy shocks rarely stay confined to energy markets. Higher fuel prices weaken household purchasing power, raise firms’ costs and complicate central-bank efforts to control inflation. The OECD expects G20 headline inflation to decline from 4.1% in 2026 to 3.6% in 2027, but warns that stronger price pressures, weaker real income growth and higher interest rates will restrain momentum.
The result is a global economy that is growing, but not comfortably. The headline numbers conceal a widening contrast between sectors and countries. Technology-related industrial production has remained strong, particularly in Asia, the United States and some European economies. Other businesses face weaker demand, expensive credit and higher operating costs.
The AI boom is doing more than creating headlines
AI has become a macroeconomic force because its expansion requires vast purchases of semiconductors, data-center equipment, electricity, construction services and networking infrastructure. That investment is supporting trade even as other parts of the economy lose momentum.
The OECD describes robust AI-related activity as one reason growth is expected to strengthen slightly in 2027. Its earlier Economic Outlook similarly identified technology investment—especially in the United States, Canada and parts of Europe—as a source of resilience.
Optimists see this as the beginning of a productivity cycle. Faster adoption could allow companies to produce more with fewer resources, lift wages in high-demand occupations and create room for central banks to ease policy without reigniting inflation. The OECD argues that stronger skills, more dynamic markets and faster technology adoption are essential to raising productivity.
The opposing view is that current growth may be overly concentrated in capital spending before the economic payoff is proven. If expected returns on AI infrastructure disappoint, investment could slow sharply. Economies heavily exposed to the technology supply chain would then face a double setback: weaker exports and less business spending.
What comes next
The next phase will depend on whether the energy shock fades faster than inflation and whether AI investment spreads beyond a small group of firms and countries. The OECD assumes energy prices will ease in line with futures markets, helping growth strengthen through 2027. That assumption remains vulnerable to renewed conflict, shipping disruptions or a larger interruption to oil supplies.
Central banks face a difficult balance. Cutting rates too quickly could revive price pressures, while keeping borrowing costs high could weaken housing, business investment and heavily indebted governments. Policymakers must also judge whether higher inflation reflects temporary energy costs or a more persistent loss of purchasing power.
Governments have a second challenge: converting technology investment into broad-based productivity rather than subsidizing a speculative boom. That means expanding electricity and digital infrastructure, improving worker training and ensuring smaller firms can access new tools. It also means managing the environmental cost of data centers and the distributional effects of automation.
The immediate outlook is therefore neither a crisis nor a clean recovery. The global economy has demonstrated that it can withstand a major shock when inventories, public support and technology investment work together. But resilience built around one fast-growing sector is not the same as balanced growth. The coming year will show whether AI becomes a durable productivity engine—or merely the latest temporary cushion.