BIS Warns Trillion-Dollar AI Capex Boom Risks Global Recession

The Bank for International Settlements has issued a stark warning about artificial intelligence investments, cautioning that the industry's unprecedented capital spending could precipitate a global recession if expected returns fail to materialize.

In its 2026 annual report, the BIS—which serves as the central bank for central banks—drew parallels between today's AI investment frenzy and historical speculative episodes including 1800s railway mania, 1920s electrification exuberance, and the 1990s dotcom boom. According to the report, first detailed by The Register, all these periods shared a common pattern: genuine technological breakthroughs that attracted capital far exceeding what commercial returns could justify, ultimately ending in investment reversals and economy-wide recessions.

The scale of hyperscaler spending

The five largest cloud providers are projected to spend more than $1 trillion on AI-related capital expenditures in 2026 alone. Amazon forecasts $200 billion in capex, Microsoft projects $190 billion, Google approximately $180 billion, Meta up to $140 billion, and Oracle is also committing substantial resources to AI infrastructure.

These commitments are outpacing the companies' earnings and free cash flow, forcing some to issue debt for additional financing. The BIS attributes this investment race partly to the perception that only a small number of players with superior technology will dominate market share, creating intense competitive pressure.

Multiple risk factors converging

The report identifies several compounding vulnerabilities. Competition is driving firms to overcommit resources to projects with uncertain returns, and as spending escalates, the net economic surplus for the tech industry declines—potentially turning negative in adverse scenarios.

Supply-side constraints present another concern. AI datacenters are straining electricity availability, exacerbating chip shortages, and creating grid connection bottlenecks. These infrastructure pressures are already affecting energy prices and input costs with potential spillover effects on inflation.

The BIS notes that temporary shortages may amplify over-investment as firms lock in future capacity through long-term contracts, further exposing them to demand disappointments. Should inflation spike or AI investment collapse, existing financial vulnerabilities could amplify macroeconomic consequences.

Financial system exposure

The rising leverage of AI companies and their growing footprint in credit markets means a shift in sentiment could trigger serious financial knock-on effects. Vulnerabilities extend throughout the supplier ecosystem, including engineering and construction contractors with comparatively weak balance sheets who remain exposed to any capex pullback by hyperscalers.

The BIS also highlighted the opacity of AI-sector financing, with corporations creating webs of private arrangements and often failing to fully disclose datacenter facility lease terms.

Why it matters

While enterprises report some employee-level efficiency gains from AI pilots, few have documented discernible productivity gains from projects deployed at scale in production environments. This disconnect between massive infrastructure investment and measurable business value creates systemic risk that extends beyond the technology sector. If the productivity promises of AI fail to materialize, the resulting financial pullback could trigger the kind of macro-financial feedback loops that characterized previous technology bubbles—but with today's more interconnected global financial system amplifying the impact.

The Register first reported these findings from the Bank for International Settlements' 2026 annual report.