The chief economist of the Bank for International Settlements (BIS) has issued a stark warning about the current surge in corporate and governmental spending on artificial intelligence (AI) technologies. According to his analysis, the rapid escalation of capital expenditures in the AI sector is being financed largely through opaque forms of borrowing, a practice that could sow the seeds of a systemic financial crisis if left unchecked. In his recent address, the BIS chief highlighted that the AI investment boom mirrors past episodes of speculative excess, notably the railway mania of the 19th century and the dot‑com bubble of the late 1990s. In each of those historical cases, investors poured massive sums of money into emerging technologies based more on optimism and hype than on solid, profit‑generating business models.
While those earlier bubbles eventually burst, the fallout was largely confined to the sectors directly involved, with broader economic repercussions being relatively limited. What distinguishes the present AI wave, however, is the scale and the manner in which financing is being sourced. The BIS official noted that many firms—ranging from start‑ups to established multinational corporations—are turning to complex debt instruments, special purpose vehicles, and off‑balance‑sheet financing arrangements to fund AI research, development, and deployment.
These financing structures are often shrouded in secrecy, making it difficult for regulators, investors, and even the companies themselves to fully assess the true level of leverage and associated risk. The chief economist warned that such opacity can mask a dangerous buildup of systemic risk.
When large swaths of the financial system are exposed to similar hidden liabilities, a shock—such as a sudden slowdown in AI‑related revenues, a regulatory clamp‑down, or a broader macro‑economic downturn—could trigger a cascade of defaults. The inter‑connectedness of modern financial markets means that distress in one segment can quickly spill over to banks, shadow banking entities, and sovereign debt holders, amplifying the impact far beyond the original AI sector.
To illustrate the potential magnitude of the problem, the BIS report cited data indicating that AI‑related capital outlays have risen by more than 150 % year‑over‑year in the past twelve months, outpacing growth in traditional IT spending. Yet, a substantial portion of that growth is being financed through instruments that lack transparent reporting standards. For example, many firms are issuing convertible bonds that can be swapped for equity under favorable conditions, thereby postponing the recognition of debt on their balance sheets.
Others are leveraging securitized assets tied to future AI revenue streams, creating a web of contingent liabilities that are difficult to trace. The BIS chief emphasized that the danger is not merely theoretical. He pointed to recent incidents where companies announced massive AI projects only to later scale them back or abandon them entirely when projected returns failed to materialize.
In several cases, those firms had already taken on significant debt to fund the initiatives, leaving them vulnerable to cash‑flow shortfalls. The resulting write‑downs and restructuring efforts have already begun to affect credit markets, with some lenders tightening conditions for AI‑related loans.
In response to these developments, the BIS has called for a coordinated policy response on multiple fronts. First, greater transparency is essential.
Regulators should require detailed disclosure of AI‑related debt, including the terms, maturity profiles, and any embedded options that could affect the true economic exposure. Standardized reporting frameworks would enable investors and supervisory authorities to monitor leverage levels more effectively and to identify concentrations of risk across institutions. Second, the BIS recommends that central banks and macro‑prudential authorities incorporate AI‑related credit growth into their risk‑assessment models. By treating AI financing as a distinct category, policymakers can better gauge whether the sector is overheating and can intervene pre‑emptively with tools such as counter‑cyclical capital buffers or targeted stress‑testing exercises.
Third, the BIS urges corporations to adopt more disciplined investment strategies. Rather than chasing market hype, firms should conduct rigorous cost‑benefit analyses, focusing on projects with clear pathways to profitability.
This approach would reduce the reliance on speculative financing and align capital allocation with sustainable, long‑term value creation. Finally, the BIS stresses the importance of international cooperation.
Because AI development is a global endeavor, cross‑border financing arrangements are common, and inconsistencies in regulatory regimes can create loopholes that obscure risk. Harmonized standards and information‑sharing agreements among major economies would help close these gaps and ensure that hidden debt does not become a blind spot for global financial stability. In conclusion, the BIS chief’s warning serves as a timely reminder that while AI promises transformative benefits across industries, the way it is being financed could undermine the very economic foundations that support its growth.
By demanding greater transparency, strengthening supervisory tools, and encouraging prudent investment behavior, policymakers can mitigate the systemic hazards posed by an opaque AI debt buildup. Failure to act, however, could leave the global financial system exposed to a new kind of bubble—one that is less visible on the surface but potentially far more disruptive when it finally bursts.