S&P Warns Hidden AI Debt May Hit Tech Giants
S&P Global warns the world's biggest tech firms may carry more hidden AI debt than reported, as over $7 trillion in AI data center spending keeps piling up.
S&P Global has warned that the world's biggest technology companies may be sitting on more AI-linked debt than investors can see. Speaking at a roundtable in London, the rating agency's digital infrastructure head said complex financing deals could be hiding big liabilities off the balance sheets of the firms leading the AI spending race.
What S&P said
Mark Habib, S&P Global's head of digital infrastructure, said hyperscalers may be using what he called "interconnected financing structures" to fund the AI buildout. In plain terms, big tech companies may be quietly guaranteeing huge data center projects or propping up smaller AI companies with their own credit, without the full cost showing up in public filings. "It is safe to assume that there is more (debt) out there that we are not seeing," Habib said, according to Reuters. He added that if a liability is not reported, it will not appear in a hyperscaler's disclosures, and if S&P is not rating the transaction, the agency may not have visibility into it either. For now, he said there was no need to change any credit ratings.
The $7 trillion spending wave
The warning lands against a staggering backdrop. S&P estimates the six largest US hyperscalers - Amazon, Microsoft, Alphabet, Oracle, Meta, and SpaceX - will spend more than $7 trillion on data centers and other AI-related capital expenditure between 2025 and 2030. The agency expects their free cash flow to stay negative until at least next year. The debt market is already showing strain. Bonds issued by AI infrastructure providers and so-called neocloud firms like Nvidia-backed CoreWeave have fallen to stressed levels in recent weeks as investors question the scale of their borrowing. A selloff in bonds linked to the AI buildout has raised the question of whether investors see more risk than the ratings suggest.
Oracle already felt the pressure
S&P pointed to an earlier warning sign. In July, the agency downgraded Oracle to BBB-, the lowest rung of investment grade, citing aggressive AI spending and an uncertain path to profitability. Habib said the biggest risk to hyperscaler ratings overall is simple: the expected profits from this enormous spending may fail to materialize on time.
Why it matters
Big tech's AI bet is the largest infrastructure spend in corporate history, and S&P is now saying the true bill could be bigger than the public numbers show. The agency is not predicting a crisis, but it is telling investors to look beyond the balance sheet. If some of this debt is hidden behind guarantees and partner deals, the first sign of trouble may only appear when a project or a partner starts to wobble.