S&P Global flags potentially hidden AI debt at hyperscalers as AI spending surges

S&P Global has raised concerns about potential hidden AI-linked debts at major technology companies, commonly referred to as hyperscalers, as they continue to invest heavily in data centers and AI infrastructure. Mark Habib, S&P Global's Head for Digital Infrastructure, stated that while there is no immediate need to change the agency's ratings, there could be undisclosed liabilities that could pose risks for these companies.

Habib explained that complex financing arrangements might not immediately appear on the balance sheets of hyperscalers, which makes it difficult to identify these hidden debts. He emphasized that if this information is not reported, it won't be visible in the disclosures made by these companies, nor will it be considered in their ratings.

The warning comes as borrowing by AI infrastructure providers and neocloud companies has been under stress. Specifically, Nvidia-backed CoreWeave has faced debt pressure due to the aggressive financing needed to fund the AI buildout. S&P Global expects the six largest US hyperscalers - Amazon, Microsoft, Alphabet, Oracle, Meta, and SpaceX - to spend over $7 trillion on data centers and other AI-related capital expenditures between 2025 and 2030.

This spending spree is also expected to keep their free cash flow negative until at least the following year, highlighting the importance of how quickly AI investments translate into revenue and profits. However, Habib said that S&P does not have an estimate for how much additional debt such interconnected financing structures could ultimately add to hyperscalers' balance sheets. Instead, the agency aims to take a holistic view of the risks involved.

Currently, S&P's biggest concern for hyperscaler ratings is whether the anticipated profits from massive AI investments will materialize. In July, the agency downgraded Oracle to BBB, the lowest rung of investment grade, citing concerns over its AI spending and an uncertain path to profitability. While S&P sees no major disconnect in bond markets that would force a rating change at the moment, Habib noted that their ratings are designed with a long-term view in mind.

A change could occur if the scale of AI spending becomes unsustainable - driven by further increases in investments by AI firms or a perception that competition in the AI sector is becoming too intense.