The margin for error in the AI funding ecosystem is narrowing, according to a recent S&P Global report, which highlights the fact that increased adoption and monetisation of the technology is needed to justify related spending plans.
In 2026, S&P Global predicts that the five largest cloud service providers (Alphabet, Amazon, Meta, Microsoft and Oracle) will spend approximately US$750 billion on capex – 38% of their revenue. Capital spending is expected to increase to US$1 trillion by 2029.
Direct debt issuance from US hyperscalers was US$115 billion in Q1 this year, up from US$70 billion across all of 2025. This activity has made the US high-tech sector the largest contributor (19%) to global nonfinancial issuance over the quarter.
Such spending has increased revenue and net income for hyperscalers in Q1, but also limited free cash flow, the report warns. It expects hyperscalers to start recovering from this in 2027/28, but added “if the expected cash flow improvements don’t materialise, stretched balance sheets could pose a risk to the ratings”.
As hyperscalers invest in other aspects of the AI ecosystem, such as data centre funding, and provide contractual support to AI model developers, interconnected financial exposure across the space increases. This circular funding is accompanied by contingent guarantees, wherein investment grade companies act as guarantors to often speculative or unrated firms.
These instances have prompted the ratings agency “to adjust hyperscalers’ debt to include obligations that behave like debt even if they’re not classified as such,” S&P Global explained.
“These types of transactions could increase future cash strains and potentially affect credit quality,” it noted, adding that each guarantee is assessed on a case-by-case basis.
S&P Global expects some hyperscalers to be better equipped to absorb their massive investments without adversely pressuring their credit metrics. Alphabet (AA+/Stable/A-1+) and Microsoft (AAA/Stable/A-1+) lead the pack with more balance sheet head room, while Oracle (BBB/Negative/A-2) lags with weaker internal cash flow and higher leverage.
Although Amazon (AA/Stable/A-1+) and Meta (AA-/Stable/–) “have seen a more pronounced erosion of financial flexibility versus historical levels as AI-driven capital spending has accelerated”, the agency expects their cash flow to strengthen as their AI investments push up revenues and EBITDA.
“The long-term credit outlook will depend on whether revenue growth can effectively outpace these massive debt-funded capital commitments.”
©Markets Media Europe 2025
