Slowing AI development could boost hyperscaler balance sheets

A Semafor review of analyst estimates finds that at current projections, as tracked by S&P Capital IQ, Amazon, Alphabet, Meta, and Microsoft are expected to spend $66 billion more on capital spending Any slowdown in AI spending would let hyperscalers catch their financial breath and ease their reliance on borrowing to fund their buildouts. A Semafor review of analyst estimates finds that at current projections, as tracked by S&P Capital IQ, Amazon, Alphabet, Meta, and Microsoft are expected to spend $66 billion more on capital spending (nearly all of it AI-related) than they earn in cash from operations over the next six quarters. Debt, stock sales, and complex deals that shift that spending off their balance sheets are filling the gap. Even a 10% pullback in capex flips that to a $74 billion surplus; a 20% pullback puts them more than $200 billion in the black. These companies might still choose to tap outside funding — “always spend other people’s money” remains a cardinal rule of corporate finance — but they wouldn’t be forced to. They could spend some time rebuilding balance sheets that have become stretched by breakneck spending or restoring stock buybacks that have largely be