[Insight & Opinion] IMF Outlook Warns of an AI Boom Retreat

Investment boom is 3.3 times the 'dot-com bubble'
Sustainability in doubt, rattling stock markets
Could it become a time bomb worse than 2008?

The International Monetary Fund's (IMF) annual World Economic Outlook, released last month, appears optimistic on the surface. Despite the shock of the war in the Middle East, it forecasts that, driven by artificial intelligence (AI), global economic growth will "continue at 3.1% this year and 3.4% next year." But a close reading of the report (WEO) reveals underlying unease and deep concern about the economic outlook.

This is especially evident in its baseline scenario. The report projects that AI-driven capital investment and productivity gains will lift global economic growth through next year, before it gradually recedes to 3.3% in 2028, 3.2% in 2029, and 3.1% in 2030. A more serious concern is that this AI-driven growth in the global economy carries diverse and complex risks.

IMF Managing Director Kristalina Georgieva also stressed the importance of managing the macroeconomic and financial stability risks that accompany the development of an AI-driven economy. IMF Chief Economist Silvana Tenreyro likewise wrote that "the global economy is resilient," but added that "this resilience is being tested." She also raised the need to address issues including excessive public debt in advanced economies, whose sustainability is in question, as well as high interest rates and inflation.

IMF Managing Director Kristalina Georgieva. Reuters-Yonhap News

Notably, the IMF report points to the possibility that its economic outlook scenario is overly optimistic in light of latent risks. It also finds that, if the AI boom collapses because of problems facing the U.S. data center investment boom, its forecasting models suggest global economic growth in 2028 could fall by as much as 1.3 percentage points. This underscores the need to take a close look at the risks embedded in the AI boom.

The first risk factor is the unprecedented scale of investment. A report by the U.S.-based Brookings Institution projected that U.S. capital spending on the AI industry from 2025 to 2032 would reach 3.63% of gross domestic product (GDP). That is 3.3 times the 1.1% recorded from 1996 to 2003, when the dot-com bubble emerged. In other words, investment spending is now at a historic high. It is also estimated that operating the AI industry built with this massive investment will require spending equivalent to 9% of U.S. GDP in 2033, raising questions about whether such spending is sustainable.

Second, market doubts and concerns are mounting over the profitability and financial risks of data center developers. The sharp drop in CoreWeave bond prices on October 14 and OpenAI's recent funding crunch are representative examples. Third, opposition from residents to data center construction in 40 U.S. states has led to the cancellation or postponement of 146 projects, prompting expectations of credit rating downgrades for data center operators. Fourth, U.S. consumer price inflation reached 3.7% last September, increasing pressure on the Federal Reserve's interest rate policy. Rate hikes would add to the burden on AI companies and could deliver a major shock to the stock market.

An adjustment in the AI investment boom is inevitable. Some, however, take the optimistic view that any correction would be limited to individual companies and would not have a devastating impact on the AI industry as a whole or the U.S. economy. But, as the Nasdaq 100's 9.1% drop in October showed, we are seeing sharp declines in U.S. stock markets almost every week amid the risks of excessive AI investment. In other words, we are already hearing thunder rumbling in the distance.

No one can know whether this will amount to a passing shower or become a major disaster in the future. But if the AI boom collapses, the integration of the global economic structure means the resulting shock could be more severe than the "2008 global financial crisis." We must therefore pay close attention to the risk of an AI boom collapse embedded in the IMF's economic outlook.

Kim Dongwon, former visiting professor at Korea University

This content was produced with the assistance of AI translation services.