AI spending drags big tech free cash flow to 2014 lows

Amazon, Alphabet, Meta and Microsoft are set to record their weakest combined free cash flow since 2014 as large-scale AI investments cut cash generation this year.

Morgan Stanley estimates that hyperscalers including Amazon, Alphabet, Meta, Microsoft and Oracle could spend about $805 billion on AI this year, up from an earlier forecast of $765 billion. The firm projects spending could reach roughly $1.1 trillion next year. Morgan Stanley notes that one year of this hyperscaler spending would be comparable to what all non-tech S&P 500 companies spent combined in the prior year.

Data compiled from sell-side analysts by Visible Alpha shows the group’s combined free cash flow could fall to about $4 billion in the third quarter, down from a roughly $45 billion quarterly average since the pandemic. Those estimates point to full-year free cash flow for the group at its weakest level since 2014, when their combined revenues were about one-seventh of today’s size.

Company-level projections highlight differences within the group. Amazon is forecast to burn about $10 billion of cash this year and has said it plans to invest $200 billion in 2026. Meta has issued about $55 billion of debt over the past six months and paused its share repurchase program; analysts expect Meta to be cash-negative in the second half of the year. Alphabet is expected to remain free-cash-flow positive for the full year but at its weakest performance in more than a decade and did not repurchase shares in the first quarter for the first time since starting its program in 2015. Microsoft is increasing capital and operating expenditures tied to AI infrastructure and services.

Analysts summarized the trade-offs the companies face: “After largely funding their investments from their income for the first few years of the AI boom, these tech giants face trade-offs more familiar to capital-intensive businesses: cutting jobs, reducing shareholder returns or borrowing to fund the build-out.”

Analysts expect the cash squeeze to ease as AI-related revenue and margin improvements materialize next year. In the near term, corporate treasuries and investors are weighing decisions on dividends, buybacks, debt issuance and workforce changes while the large-scale AI infrastructure is built.