"The capex intensity is unlike anything we've seen before," one analyst noted during late July earnings calls. Amazon, Alphabet, Microsoft, Meta, and Oracle just guided for combined capital spending exceeding $725 billion in 2026, a stunning 77% jump from 2025. Most of it flows straight into data centers, custom chips, and infrastructure to feed the AI beast. The result is jarring: companies that once reliably printed cash are now bleeding money on a free cash flow basis.
Alphabet reported its first negative free cash flow quarter since going public in 2004, losing roughly $5.9 billion in Q2 2026 alone, with quarterly capex hitting $44.9 billion. Amazon's trailing twelve-month free cash flow sits at negative $7.6 billion. Meta's cash generation collapsed 91% year-over-year. But Oracle's numbers might hit hardest, burning through nearly twice what it generated from operations. The company is now raising between $45 billion and $50 billion through debt and equity just to fund the spending spree. Wall Street watched these firms report strong profit growth and healthy cloud revenue throughout earnings season, then fixated entirely on the cash flow deterioration.
The scale keeps expanding. Analysts project AI-related capex could hit $765 billion in 2026 and balloon to $1.2 trillion by 2027. Aggregate hyperscaler spending on compute, data centers, and power from 2025 to 2030 lands around $5.3 trillion, with 2026 to 2031 estimates stretching to $7.6 trillion. Rising memory chip costs pile pressure on top of construction and equipment bills that are already staggering. This arms race kicked off in late 2023 when generative AI forced every major cloud provider to accelerate buildouts simultaneously, all chasing customer demand for AI compute capacity that seemed to have no ceiling.
This article is for informational purposes and should not be construed as financial advice. AI capex trends affect market dynamics broadly, but individual investment decisions require independent analysis.
