IMF Warns AI Infrastructure Spending Risks Triggering Financial Crisis

The latest Global Financial Stability Report highlights that while hyperscalers—including Microsoft, Meta, Amazon, Alphabet, and Oracle—possess strong core earnings, their free cash flows are eroding under the weight of massive AI investment requirements. This shift has forced firms to rely heavily on debt financing and complex circular agreements, where companies fund each other’s projects in exchange for reciprocal product purchases. The IMF warns that this creates "wrong-way risk," where investments depreciate simultaneously as revenue streams falter.

Investment expert Bret Jensen notes that the five largest hyperscalers now carry a combined $1.3 trillion in on-balance-sheet debt, with an additional $3 trillion in off-balance-sheet allocations. According to Jensen, debt markets are already showing signs of strain, signaling potential default risks for firms like Oracle. Much of this infrastructure supports "unprofitable" and cash-strapped labs like OpenAI and Anthropic.

Economist Dean Baker argues that current AI investment is propping up the U.S. economy on the fragile hope of unprecedented productivity gains. However, data shows productivity growth remains at its long-term trend pace. Without the promised surge in efficiency, experts fear a market correction could lead to a guaranteed recession.