The International Monetary Fund published a technical note assessing the financing structure and financial stability implications of the artificial intelligence investment boom. Current risks appear manageable because the expansion remains anchored by profitable incumbent firms with strong balance sheets. However, rising capital intensity, external financing, market concentration and interconnected commercial arrangements could amplify shocks if AI monetization or productivity gains disappoint. Total AI related capital expenditure is projected to exceed USD 4 trillion by 2029, with hyperscalers accounting for nearly 75%. Financial strength varies sharply across the AI ecosystem. Hyperscalers and chip developers have strong profitability and contained leverage, while data centers and graphics processing unit cloud providers are more constrained by debt, weak profitability and high capital expenditure. Funding is shifting from internal cash flow toward bonds, private credit and securitization. The note also examines 14 disclosed circular financing deals in which equity investments of about USD 300 billion supported USD 1.4 trillion of future revenue commitments, while nine deals were followed by more than USD 650 billion of additional debt financing. Faster technological obsolescence, correlated equity positioning and opaque exposures could intensify losses and transmit stress across firms and funding markets. The note calls for greater transparency and data collection, monitoring of interconnected and cross-border exposures, scenario-based risk assessments, and closer attention to leverage and risk-taking by banks and nonbank financial institutions.
International Monetary Fund technical note finds AI financial stability risks manageable as leverage and interconnectedness rise
An International Monetary Fund technical note finds that AI related financial stability risks remain manageable but could grow as investment relies more heavily on external and interconnected financing. Vulnerabilities are concentrated among infrastructure and cloud providers, while circular financing, technological obsolescence and correlated markets could amplify shocks. The note highlights the need for better exposure data, scenario analysis and monitoring of leverage across banks and nonbank financial institutions.