AI Sector Debt Surge Creates Systemic Financial Risk

The artificial intelligence sector is undergoing extreme financialization characterized by circular financing deals and significant debt accumulation. CoreWeave, OpenAI, and Nvidia have entered interlocking arrangements where investments are used to fund hardware purchases that are subsequently rented back to the investors. This leverage includes the use of GPUs as loan collateral, a practice analysts compare to the triggers of the 2008 financial crisis.

CoreWeave has accrued 14 billion dollars in debt and faces 34 billion dollars in upcoming lease payments, with Microsoft providing up to 70 percent of its revenue. Simultaneously, OpenAI projects losses of at least 15 billion dollars this year despite signing massive computing deals with Amazon, Oracle, and CoreWeave. Nvidia has expanded its reach through over 50 deals this year, including a 100 billion dollar investment in OpenAI.

To manage these costs, tech giants including Meta, Google, and Amazon are using special-purpose vehicles to keep billions in data center debt off their balance sheets. For instance, Meta partnered with Blue Owl Capital for a 27 billion dollar data center project via such a vehicle. This trend is further supported by the Trump administration, which issued an executive order loosening regulations to allow 401(k) holders to invest in alternative assets like private credit. Experts warn that a failure to generate short-term profits could cause chip values to collapse, triggering widespread defaults across private-credit firms and financial institutions.