Almost four years into the AI boom, the world's biggest technology companies are still promising grand returns — but they are burning through cash to get there. Amazon, Alphabet and Tesla all posted negative free cash flow in the latest quarter, while Meta's cash generation plunged 91% from a year earlier, according to a CNBC analysis of this earnings season.
Goldman Sachs projects megacap AI spending will reach $765 billion this year and approach $1.2 trillion in 2027. Amazon raised its 2026 capital spending forecast to $220 billion — the highest among the hyperscalers — and reported negative free cash flow of $7.6 billion over the trailing twelve months. Alphabet said free cash flow turned negative for the first time on record, a stunning development for one of the most profitable companies on the planet.
A key reason costs are rising faster than expected is the memory crunch: AI processors depend on memory supplied by a small set of vendors, and prices are soaring. Tesla CEO Elon Musk called memory pricing "insane" and thanked Micron for a "very significant allocation on reasonable terms," while Amazon CEO Andy Jassy said inflated memory prices drove his company's capex guidance higher.
Investor reactions diverged sharply. Tesla and Alphabet sank after turning cash-flow negative; Meta fell after a weak forecast; Microsoft, by contrast, had its best day since 2008 after posting a 31% profit jump alongside higher capex guidance. Apple — the most conservative spender — flagged "supply constraints" as memory shortages pressure its products.
Wall Street is increasingly questioning whether the buildout will pay off. JPMorgan strategists asked whether markets are betting on "one big AI trade," and analysts say investors are trying to separate AI winners from losers as balance sheets distort. With Nvidia reporting on Aug. 26, the next test of the AI trade is already on the calendar.