In a speech in Singapore on October 7, ahead of the IMF and World Bank Annual Meetings in Bangkok, IMF Managing Director Kristalina Georgieva described an awkward timing problem. Spending on AI infrastructure is lifting demand and corporate earnings now. The broader productivity gains that could justify that spending will take longer, and countries outside the technology supply chain may receive much less of the near-term boost. Reuters and AP reported her remarks; the IMF's event page identifies the speech and subsequent discussion with Singapore President Tharman Shanmugaratnam.

The big change

  • What changed: The IMF chief put the AI buildout beside energy prices and public debt as a current macroeconomic policy problem. She warned that concentrated investment could transmit an earnings disappointment through financial markets.
  • Why it matters: Investment can raise growth before the promised productivity gains appear. Countries making chips and supplying AI infrastructure may benefit first, while energy demand and higher financing costs can reach a wider set of economies.
  • What to watch: The IMF's forthcoming World Economic Outlook, evidence of productivity gains beyond construction and technology firms, and whether governments improve worker skills, energy security and fiscal resilience. Georgieva's growth upside remains a projection.

An investment boom with a delayed payoff

Georgieva said AI investment as a share of GDP is likely to surpass the historical scale of railroads, electricity grids or telecommunications networks, according to both news reports. That is a comparison about expected investment intensity, not evidence that AI has already delivered comparable economic returns. AP reported her warning that large data-center investments and their eventual benefits arrive at different times. Reuters reported her concern that concentrated ownership, debt and high expectations for AI companies could turn disappointing earnings into a wider market shock.

She also cited IMF research suggesting that AI, if used well, could add about half a percentage point to annual global growth, Reuters reported. That is potential additional growth, not a measured contribution in 2026. The IMF's July World Economic Outlook update projected global growth of 3.0% for 2026 and described AI demand as a lift for economies tied to the technology value chain. The October forecast had not been released at the time of the speech, and Reuters reported that Georgieva did not say whether the global 2026 figure would change.

There is a narrower observed figure. The IMF's 2026 annual report estimated that technology investment related to AI added 0.5 percentage point to US GDP growth in 2025. It is an estimate for one country's investment contribution in one year. It does not show a half-point annual gain in world productivity or establish the future return on today's AI spending.

Why the gains may land unevenly

The immediate demand boost follows factories, chips, data centers and related supply chains. Georgieva said China, India, Japan, South Korea, Taiwan and other economies with strong technology sectors are benefiting while many others are being bypassed, AP reported. The IMF's July outlook likewise says the technology upcycle helps countries integrated into the global value chain, while the war-related energy shock weighs on importers and vulnerable economies. That contrast describes the IMF's current assessment; it does not measure AI's eventual effect on every country's income.

The IMF's AI Preparedness Index offers a way to examine a different part of the question: capacity to adopt AI. It covers digital infrastructure, human capital and labor policy, innovation and integration, and legal and regulatory conditions across 174 economies. The IMF cautions that it is an indicative measure, partly based on perceptions, and should not be treated as a ranking of realized AI benefits. A country can score well on readiness without having demonstrated an economy-wide productivity gain.

Energy tightens the link between the boom and the rest of Georgieva's warning. Reuters described her account of a negative energy supply shock from Middle East conflict meeting the demand added by AI construction. AP reported that she pointed to rising AI energy demand alongside higher prices for other essentials. The speech did not establish how much of the current global price rise is attributable to AI. It did put power supply and energy security into the economic calculation for AI investment.

Policy while returns remain uncertain

Georgieva urged governments to rebuild fiscal strength, protect vulnerable people, improve worker skills, make it easier to start and close businesses, strengthen energy security and set AI regulatory guardrails, according to Reuters and AP. Reuters also reported her call for credible medium-term fiscal plans in high-debt advanced economies and a prudently hawkish monetary stance in many countries. These are her policy recommendations, not measures adopted by the member governments meeting in Bangkok.

For economic policymakers, the test is whether AI spending turns into broader productivity and income gains before debt, energy costs or an earnings reversal impose wider costs. For readers, the near-term evidence is investment and its uneven demand effects. The size and distribution of any lasting AI growth dividend remain open.

Sources & further reading