BIS Chief Warns AI Spending Boom Relies on Opaque Debt, Risks Financial Stability

Bank for International Settlements head Pablo Hernandez warned that the rapid growth in artificial intelligence investment could threaten broader financial stability if companies fail to generate returns matching investor expectations. He compared the current AI spending surge to historical episodes including the 1830s canal mania, 1840s British railway boom, 1920s electrification wave and the late 1990s dot com bubble, noting that all eventually drew in more capital than returns could justify.

Debt Financing Raises Systemic Concerns

Hernandez noted that major AI companies’ capital spending is increasingly outpacing their cash flow, relying instead on debt and private credit as firms compete in an escalating investment race. He warned that financing arrangements linking chipmakers, cloud providers and AI firms can be difficult to value, leaving the broader financial system exposed if profit expectations fall short. He pointed to projections that global AI-related investment could grow from roughly $500 billion today to between $3 trillion and $4 trillion by 2030, with the five largest tech companies alone planning to spend over a trillion dollars on AI projects between 2025 and 2026.

Global Ripple Effects Possible if Expectations Disappoint

Given the outsized weight of US equities in global markets, Hernandez said a correction in concentrated AI stocks could have effects extending well beyond American borders, potentially impacting household spending and contributing to asset bubbles in countries benefiting from AI related export gains. While acknowledging genuine productivity benefits from AI in areas like coding and professional writing, he stopped short of declaring the current boom unsustainable, saying only that its scale and pace warrant caution. He added that AI does not change central banks’ core mandate, though it may complicate efforts to interpret and monitor the broader economy.