US AI spending fuels hidden debt surge

Analysts continue to highlight the scale of off‑balance‑sheet liabilities tied to AI investments by the five U.S. tech giants—Alphabet, Microsoft, Amazon, Meta and Oracle. By late July 2026 the hidden obligations had risen to about $1.65 trillion, an eight‑fold increase over four years, dwarfing the firms’ reported $1.35 trillion debt. A Nikkei analysis released on 23 July noted that Meta alone may carry $420 billion of such liability, underscoring the concentration of risk.

The same day, U.S. markets saw a single‑day erosion of more than $800 billion in market value across the “Magnificent Seven” tech firms. Tesla led the sell‑off with a 14 % drop, wiping out $226 billion, while Alphabet lost $290 billion and Amazon, Meta and Microsoft each lost billions. The plunge was attributed to doubts over massive AI‑related capital expenditures and the profitability of those investments.

Financing structures—leases, special‑purpose entities and large loans such as the $27 billion financing of Meta’s Louisiana data centre—keep these exposures off traditional balance sheets, prompting fears that a slowdown in AI demand could raise financing costs, leave data centres under‑utilised, and trigger a loss of investor confidence. Commentators draw parallels to the Enron collapse. Market reactions remain uneven; Apple’s stock fell about 8 % while Microsoft’s dropped roughly 20 %. Investors now focus on upcoming quarterly reports to see whether AI‑driven spending translates into sustainable revenue or further debt pressure.