S&P Global warns hyperscalers may carry more AI-linked debt than disclosed

S&P Global estimates major technology companies may have more AI-linked liabilities than investors can currently see. Mark Habib, its digital-infrastructure chief, pointed to financing in which hyperscalers effectively guarantee data-centre projects or smaller AI companies.

S&P expects six US hyperscalers to spend more than $7 trillion on data centres and other AI capital expenditure between 2025 and 2030, with negative free cash flow until at least next year.

Habib said S&P has no estimate for extra debt such structures could add and sees no immediate ratings change.

The warning is about disclosure and credit visibility, not an allegation of wrongdoing. Investors may need to examine guarantees and project entities alongside reported debt.

S&P’s concern is particularly relevant as infrastructure spending increasingly involves partnerships, leases and special-purpose arrangements whose risks may not appear as ordinary corporate bonds.

Image: S&P Global Ratings press-release page (official page's declared media asset).

Sources

• Reuters