The IMF's World Economic Outlook is the most widely cited economic forecast in the world, and the October 2026 edition carries a story that looks, at first reading, relatively reassuring: global growth of 3.1% in 2026, rising to 3.4% in 2027, maintained against the headwinds of a significant energy price shock and continuing geoeconomic fragmentation. The mechanism the IMF identifies as responsible for holding the global forecast up is artificial intelligence: AI investment and AI-related trade are providing a positive offset to the negative forces that would otherwise pull global growth lower. This is a significant finding. It is also a finding that requires immediate qualification, because the AI economy that is sustaining the global aggregate is not distributed in a way that reaches the majority of the world's economies.

The IMF's own framing of the distribution question is precise: AI supply chains are concentrated in relatively few countries, sectors and firms. This is not a qualification that appears in small print at the back of the outlook. It is a central structural observation about the nature of the growth that AI is generating. The investment flowing into AI infrastructure -- data centres, semiconductor fabrication, training compute, power systems -- is going overwhelmingly to a small number of countries and a small number of companies within those countries. The trade flows being driven by AI are similarly concentrated. When the IMF says AI is counteracting the drag from the energy shock, it is describing a mechanism that operates primarily within and between a small number of advanced economies and a handful of strategically positioned emerging markets. For most developing economies, the AI contribution to global growth is statistical background rather than operational reality.

The most analytically useful way to read the WEO finding is through a three-category lens that the IMF's own research implies but does not always make explicit.

Most of Sub-Saharan Africa, much of South Asia outside India, and large parts of Latin America fall into the second or third category. They are in the global economy that the IMF is measuring. They are not in the economy that is doing the work of counteracting the energy shock.

The second structural observation in the IMF's October 2026 WEO is that high borrowing costs are reducing governments' fiscal room to respond to the situation the first observation describes. The irony is precise: the period in which strategic AI investment is most important for long-term economic positioning is also the period in which the cost of the borrowing that would fund that investment is at its highest for much of the developing world. Countries that entered the current cycle with high debt-to-GDP ratios and external debt denominated in hard currencies are facing debt service costs that constrain their ability to direct public resources toward the infrastructure, skills and industrial policy that would improve their position in the AI value chain.

This is not a temporary problem that resolves when interest rates fall. The concentration of AI investment in advanced economies is self-reinforcing: early movers attract talent, build capacity, generate intellectual property and set the technical standards that future entrants will be required to adopt. The window in which strategic entry into AI infrastructure production is possible may be shorter than the window in which most developing economies will have the fiscal capacity to make that entry. The IMF does not say this explicitly. The structure of its own data implies it.

Global growth of 3.1% in 2026 is approximately where it was before the pandemic period, and its projection to 3.4% in 2027 is not dramatically below historical averages. In that sense the WEO carries a message of relative stability rather than crisis. The stability is real. It is also, as UNCTAD's Trade and Development Report published the previous week also found, increasingly organised around a smaller number of economies and sectors than the aggregate suggests. The global growth figure is an average. Averages average. A 3.1% global figure that reflects 7% growth in India, 6%+ growth in several Central Asian economies and 4.5% growth in Sub-Saharan Africa alongside much lower or negative growth in economies facing fiscal crisis, commodity price volatility and climate disruption is not the same economic situation as a 3.1% figure distributed more evenly across the world economy.

The IMF identifies AI as the primary positive force counteracting the energy shock and geoeconomic fragmentation in the October 2026 World Economic Outlook. The AI supply chains that produce this positive force are concentrated in a small number of countries, sectors and firms. High borrowing costs reduce the fiscal capacity of developing economies to enter those supply chains. At what point does a technology that sustains the global growth figure while concentrating its gains in a narrow geography cease to function as a development opportunity for the majority of the world economy, and become instead a new structural mechanism for divergence?