Economists and central bankers convened at the European Central Bank's annual symposium in Sintra, Portugal, are expressing growing concerns about the financial stability risks associated with the artificial intelligence boom. While acknowledging AI's potential to significantly boost productivity and foster economic growth, they highlighted several worrying trends. These include the substantial increase in debt issuance by AI hyperscalers, the rising leverage used by investors to speculate on AI companies, and the potential for widespread job displacement if AI technologies replace human labor.
Tobias Adrian, director of the monetary and capital markets department at the International Monetary Fund, specifically voiced concerns about the leverage present on both the borrower and investor sides of the AI market, deeming it "very worrisome for financial stability." The Bank for International Settlements (BIS) echoed these sentiments, warning that the surge in AI spending could reverse, potentially pushing some economies into recession. Bank of Canada Governor Tiff Macklem drew parallels to the dot-com bubble, noting that while AI's potential is immense, current stock valuations for AI-linked companies appear stretched, suggesting the market might be getting ahead of itself.
Apollo Global Management's chief economist, Torsten Slok, outlined two scenarios, both with risks. If AI is wildly successful, it could lead to job losses and reduced consumer spending, potentially triggering a recession. Conversely, if AI fails to deliver on its productivity promises, the current substantial investments, like the $1 trillion committed by the five largest hyperscalers for 2025 and 2026, could prove problematic as returns may not materialize. The dominance of AI companies in global investor portfolios is also a concern; the 10 largest companies in the S&P 500, many tied to AI, now constitute approximately 40% of the index.
Further concerns include "circular financing," where chipmakers and cloud giants invest in AI labs that then commit to buying their products, effectively recycling money. Much of this funding flows through less regulated channels like hedge funds and private credit vehicles. Private credit funds alone originated over $40 billion in loans to AI-related companies in 2025. Additionally, the Bank of England's Sarah Breeden highlighted cybersecurity risks, noting that new AI models can uncover system vulnerabilities, and delayed patching could leave financial institutions exposed to crippling cyberattacks, as warned by ECB's Isabel Schnabel. The build-out costs could reach $8 trillion over the next six years, with potential for a "third wave" of inflation as chip manufacturers prioritize high-margin AI components, impacting consumer electronics pricing.
Policymakers, including BIS chief representative Zhang Tao and Financial Post contributor Hernández de Cos, emphasized the urgency of addressing these challenges, advocating for price stability, strengthened financial stability, sound monetary and fiscal foundations, and reforms for sustainable growth. They warned that delaying action would only increase the costs and complexities of necessary adjustments in the future.
