The data center boom is drawing massive financing volumes and emerging as a risk to monitor for the broader financial system. In its Financial Stability Report published on November 13, the Federal Reserve warned that debt tied to advanced computing infrastructure surpassed $680 billion in the first ten months of 2026, according to Fed data. The same report found that hedge fund leverage, particularly among the largest funds, has reached its highest level since the Fed began collecting the data in 2013. Financial Stability Report
The Fed is not forecasting a crisis and is not announcing new regulatory measures. The report examines vulnerabilities and potential stress scenarios: the expansion of AI infrastructure, elevated asset valuations, and the interconnections between companies and their financiers are all areas demanding close attention.
Advanced Computing Debt Is Accelerating Fast
According to the Fed’s Financial Stability Report, debt issuance tied to advanced computing reached $680 billion by October 2026, roughly five times the total issued across all of 2023. These financings now represent nearly one-third of non-financial investment-grade corporate issuances, up from less than 6% three years ago.
That figure covers debt connected to advanced computing infrastructure specifically, not all AI-related spending. Behind the numbers are data centers, servers, accelerators, and other investments needed to supply raw computing power. These are capital-intensive, long-horizon projects in a sector where future demand and revenues remain genuinely difficult to forecast.
The Fed draws a parallel with the late-1990s broadband telecommunications investment cycle. Back then, infrastructure buildout was fueled by sweeping growth expectations. The comparison doesn’t mean history will repeat itself. It does highlight how quickly an investment wave can translate into broad, interconnected financial exposures.
A Web of Debt, Guarantees, and Contracts
Risk cannot be measured purely by looking at loans recorded on corporate balance sheets. According to the Fed, outstanding debt, guarantees, and not-yet-activated leases tied to this infrastructure could collectively exceed $3 trillion. That estimate encompasses commitments of varying types and should not be read as already-disbursed debt.
Many of these projects involve large technology operators, hyperscalers, data center developers, private financiers, and special-purpose vehicles created to support individual investments. Long-term contracts, guarantees, leases, and equity stakes can link entities that, viewed in isolation, appear to carry relatively modest exposures.
That web of connections can cut two ways. If difficulties emerge, large companies involved may absorb some losses and limit contagion. But their actual commitments could prove far larger than balance-sheet liabilities alone suggest. The Fed therefore flags that the relationships among corporations, credit markets, private lenders. Financing vehicles deserve close monitoring.
The issue connects directly to the rapid expansion of data centers and the race to build infrastructure capable of running large AI models. Sector growth depends on a combination of capital, chips, energy, and network access. If any of those inputs becomes scarce or significantly more expensive, project economics can shift quickly.
Hedge Fund Leverage Stays Under Scrutiny
The report also flags that hedge fund financial leverage has reached its highest level since the Fed began collecting data in 2013, with the increase concentrated primarily in the largest funds. High leverage can amplify returns when markets move as expected, but it can equally accelerate losses and force investors into asset sales when prices fall.
The central bank’s concern centers on the potential combination of elevated valuations with debt-financed positions. If expectations around AI revenues were to weaken, a market correction could make refinancing some projects harder and reduce credit availability. The Fed notes that AI investment contributed significantly to U.S. economic growth in 2026, which means a sharp slowdown could carry broader consequences for markets and the wider economy.
The report does not, however, describe the banking system as being in crisis. It distinguishes between the overall soundness of banks and the vulnerabilities that can build up elsewhere, particularly in the interconnections among private finance, large technology companies, and AI infrastructure. This is a risk assessment, not a firm prediction of losses or contagion.
AI Becomes a Central Financial Risk
In a Fed survey conducted between September and October, 87% of respondents identified artificial intelligence as a material risk to financial stability, up from 50% in the May survey. These are participant opinions, not official probabilities assigned by the central bank to a crisis scenario. Concerns cited include cyberattacks, complex AI-related financing structures, and the possibility that an earnings slowdown triggers a market correction.
What’s new is that AI now features not only as a technology capable of transforming businesses and productivity, but as a phenomenon with direct consequences for credit markets. Capital is financing the construction of infrastructure that will need to be used and monetized over the years ahead. If demand, service pricing, or energy availability fall short of expectations, that pressure could propagate along the entire financing chain.
For now, the Fed’s message is one of vigilance. AI investment is supporting economic growth, but the scale of the debt and the complexity of the financial relationships make it more important than ever to understand who is financing these projects, what guarantees are involved, and where losses would fall if growth disappoints. Investors and market participants should watch the Fed’s next Financial Stability Report closely, as any shift in tone on AI-related exposures could signal changing conditions in both credit markets and equity valuations.
