AI Infrastructure Spending Drives Massive Debt Surge and Strains Credit Markets

The enormous cost of building the physical infrastructure behind artificial intelligence is pushing companies deeper into debt and creating new pressures in credit markets, according to recent reporting from Bloomberg and Fast Company. As tech giants and AI-focused firms race to build data centers, buy chips, and expand power capacity, Wall Street lenders are increasingly involved in financing deals that can run into the tens or even hundreds of billions of dollars. That borrowing spree is raising concerns about how much risk is being shifted away from the biggest technology companies and into the broader financial system.

Bloomberg reported on May 23 that the surge in “hyperscaler” spending is generating a boom in credit derivatives, the contracts banks and investors use to hedge or transfer credit risk. The article describes how banks that want to keep doing business with major cloud and AI companies are having to trade more of these instruments as the scale of borrowing grows. In plain terms, the more money these companies need, the more complicated the financing becomes — and the more financial intermediaries are needed to absorb or spread the risk.

Fast Company, in a separate piece on Oracle and the AI boom’s “hidden debt bomb,” focused on how Oracle has become a key example of the strain. Oracle has positioned itself as a major infrastructure partner for OpenAI and other AI players, but that strategy appears to be coming with a heavy financing burden. The article says the company’s ambitions are tied to massive data center projects and large outside funding needs, including estimates that AI-related spending across the industry could reach about $3 trillion by 2028. Analysts cited in the story suggest that only about half of that sum could come from companies’ own cash flow, leaving a huge amount to be funded through banks, bond markets, private credit and other sources.

The concern is not just the size of the numbers, but where the debt sits. As Fast Company reported, some of the financing is being structured through special-purpose vehicles, or SPVs, and private-credit funds, which can keep borrowing off a company’s main balance sheet. That may make the companies look less leveraged on paper, but it also means more of the risk can end up in less transparent corners of the market. The piece argues that this “hidden” financing structure is helping power the AI buildout while also making it harder for outsiders to see how much debt is accumulating.

Oracle’s role matters because it sits at the center of several major AI infrastructure deals and has committed to an aggressive expansion tied to the industry’s growth. According to the Fast Company report, analysts at Morgan Stanley estimate Oracle may need an additional $100 billion or more through 2027 and early 2028 to keep up with its commitments. That has prompted close attention from lenders, investors and competitors, especially as other large technology companies also compete to secure data center capacity and computing power.

Taken together, the two reports suggest that the AI boom is becoming as much a financing story as a technology story. The winners are not only the companies building models, but also the banks, credit funds and structuring firms that make the spending possible. But the faster the industry expands, the more pressure it places on lenders, and the more important it becomes to understand who is carrying the debt if the boom slows.