Amazon has raised around $100 billion in the bond market this year as Big Tech hyperscalers pile debt into artificial intelligence infrastructure, and S&P says the credit quality of those companies is gradually weakening.
The money is flowing into data centers and other AI-related infrastructure, with the company saying the enormous sums being spent by Big Tech hyperscalers are in the trillions of dollars, much of it borrowed and some invested in riskier businesses. The spending is helping fuel U.S. economic growth overall. It’s also putting the pristine credit ratings of the hyperscalers at risk.
Who Pays for the Boom
S&P analysts wrote, “The credit quality of hyperscalers is gradually weakening.” They added, “Every time we take a deep dive into this sector, we find that capex is rising faster than we anticipated, financings are becoming more complicated and less transparent and that returns on investment will take years to realize.” That’s the language of a system built on leverage, opacity and delay. The profits are promised later. The debt lands now.
S&P forecast that the top six hyperscalers — Amazon, Microsoft, Alphabet, Oracle, SpaceX and Meta — will spend more than $7 trillion on data centers and AI capex through 2030. That scale alone shows where the power sits: in a handful of corporate giants deciding how much capital gets poured into the machine, while everyone else lives with the consequences.
Naveen Sarma, an analyst at S&P who coauthored the report, said the risks for the hyperscalers are not “existential.” He said most of the companies, with the exception of Oracle, are highly rated borrowers with strong cash-flowing businesses. But he said the established players are lending their reputations and credit ratings to smaller and untested companies, including neoclouds, data center operators and giant startups such as Anthropic and OpenAI. That’s the old trick: the big names carry the weight, and the smaller players get pulled into the orbit.
Sarma said, “When we get together and talk about where the risks are from AI, it's these smaller companies. It's municipalities, banking on taxes from data centers and spending lots of money on infrastructure. It's utility companies building power plants. All of these ancillary things.” The costs don’t stop at the boardroom. They spill outward, into local budgets, public infrastructure and utility systems.
The Web Gets Tighter
The report said those ancillary businesses are driving a lot of economic growth, but that there are “Rumsfeldian” unknowns in the sector. The analysts cited the possibility that cheap, open-source models could undercut Anthropic and OpenAI and reduce the return to investors, uncertainty over when the massive investments will pay off, and questions about how much circular financing is inflating revenue across the industry. The whole setup depends on money moving in circles and confidence staying intact.
Much of the money companies are spending on AI is coming from debt, with Amazon alone having raised around $100 billion in the bond market this year. The hyperscalers are also backing borrowing by other players, using their strong credit ratings to help less well-known and riskier companies borrow at lower interest rates. That support can include agreeing to sign a lease in the future, backstopping a loan or buying chips. The big companies are also taking large stakes in OpenAI and Anthropic.
What the Ratings Can’t Hide
Richard de Chazal of William Blair Equity Research wrote that AI earnings are becoming more circular. He said, “If end-demand disappoints, these firms could be hit twice: first through slower revenue growth and then through lower valuations on their AI-related investments.” That’s the kind of double exposure that comes from building an economy around speculation, debt and interlocking bets.
The S&P analysts said, “The scale of overlap and interconnectedness is vast,” and warned that “In a downturn even the best capitalized and most profitable firms may incur substantial pain.” The people at the top call it growth. The report reads more like a warning label on a machine that’s already been set in motion, with municipalities, utilities and smaller firms left carrying pieces of the risk while the hyperscalers keep expanding the apparatus.