In a recent briefing, the governor of the Bank for International Settlements (BIS) sounded the alarm over the rapidly escalating competition among corporations and governments to pour capital into artificial intelligence (AI) technologies. While the promise of AI is undeniable—offering efficiencies, new products, and transformative capabilities—the manner in which the funding is being marshaled raises serious concerns about financial stability.
The BIS chief, Pablo Hernandez, highlighted that much of the current surge in AI‑related capital expenditure (capex) is being financed through opaque forms of debt. Unlike traditional loans that are transparently disclosed on balance sheets, many of the new financing arrangements are structured as hybrid instruments, special purpose vehicles, or off‑balance‑sheet commitments. This lack of clarity makes it difficult for regulators, investors, and even the borrowing firms themselves to gauge the true level of leverage being accumulated. Hernandez drew a historical analogy to two well‑known bubbles: the railway mania of the 19th century and the dot‑com boom of the late 1990s.
In both cases, optimism about future profits drove investors to fund projects that, in hindsight, were often over‑valued or fundamentally unsound. The railway era saw massive amounts of capital channeled into rail networks that could not generate the expected returns, leading to widespread bankruptcies and a contraction in credit.
The dot‑com bubble, similarly, was fueled by speculative valuations of internet companies that lacked sustainable business models, culminating in a dramatic market correction in the early 2000s. The BIS chief warned that the AI spending surge bears a striking resemblance to those past episodes. Companies are committing billions of dollars to AI research, development, and deployment based largely on hype and the fear of being left behind, rather than on rigorous assessments of projected cash flows. This behavior is especially pronounced in sectors where AI is seen as a strategic imperative—such as finance, healthcare, and manufacturing—prompting firms to take on debt that may not be fully repaid if the anticipated revenue streams fail to materialize.
One of the core risks identified is the potential for a systemic shock if a significant portion of these AI‑related debt defaults simultaneously. Because much of the financing is shrouded in complex structures, a default could propagate through the financial system in unforeseen ways, affecting banks, shadow lenders, and even sovereign balance sheets that have indirectly supported the AI investment wave.
The opacity also hampers the ability of central banks and supervisory authorities to monitor aggregate leverage levels, making pre‑emptive policy action more challenging. To mitigate these dangers, Hernandez called for greater transparency in how AI projects are funded. He urged firms to disclose the full extent of their AI‑related liabilities, including contingent obligations and off‑balance‑sheet exposures.
Moreover, he advocated for standardized reporting frameworks that would allow regulators to compare leverage ratios across industries and jurisdictions. Such measures would help identify pockets of excessive risk before they snowball into a broader crisis. In addition to transparency, the BIS chief emphasized the importance of aligning AI investment with realistic profitability forecasts.
He suggested that firms adopt more disciplined capital budgeting practices, incorporating stress‑testing scenarios that account for slower adoption rates, regulatory hurdles, and potential technology setbacks. By doing so, companies can better gauge whether the debt they are assuming is justified by credible future cash flows.
The warning comes at a time when governments worldwide are also stepping up their own AI initiatives, often providing subsidies, tax incentives, or direct funding to stimulate domestic AI ecosystems. While public support can accelerate innovation, it also adds another layer of complexity to the debt picture, especially if public funds are channeled through indirect mechanisms that lack full accounting transparency.
Hernandez’s remarks have sparked a broader debate among policymakers, industry leaders, and academic experts about the appropriate balance between fostering AI innovation and safeguarding financial stability. Some argue that the benefits of AI—ranging from productivity gains to new revenue streams—justify a degree of risk-taking, while others contend that unchecked leverage could undermine the very economic growth that AI aims to boost. In response, several central banks are reportedly reviewing their supervisory frameworks to incorporate AI‑related credit exposures. The European Central Bank, for instance, has signaled an intention to issue guidance on the reporting of AI‑driven investments, while the U.S.
Federal Reserve is exploring ways to integrate technology‑risk assessments into its stress‑testing regime. Ultimately, the BIS chief’s cautionary message underscores a timeless lesson from financial history: enthusiasm for transformative technologies must be tempered by prudent risk management. As AI continues to reshape industries and societies, ensuring that the financing behind it is transparent, sustainable, and grounded in realistic expectations will be essential to avoid repeating the costly mistakes of past bubbles. The hope is that, by heeding these warnings now, policymakers and business leaders can steer the AI revolution toward a path that delivers lasting value without endangering the stability of the global financial system.