Google, Amazon, Microsoft, and Meta together plan to spend roughly $725 billion on capital expenditures in 2026, a 77% jump from 2025's already-record $410 billion, almost entirely driven by AI data center buildout — but the stock market's reaction to that spending has split sharply by company, proving that "we're spending more on AI" alone doesn't tell investors what they actually want to know. This is reporting on disclosed earnings results and stock reactions, not investment advice.
The Number: $725 Billion and Climbing
The scale here is genuinely historic. Goldman Sachs now projects a combined $5.3 trillion in capital expenditures across Meta, Microsoft, Amazon, and Alphabet from fiscal 2025 through fiscal 2030, up from a $4.5 trillion estimate before this year's earnings season began, with a broader industry-wide baseline estimate reaching $7.6 trillion between 2026 and 2031 across compute, data centers, and power. As The Futurum Group's Daniel Newman put it after one particularly strong earnings night in April 2026, "the 'AI capex is speculative' narrative is dead" — the companies are unmistakably committing real capital at a scale bigger than the GDP of many mid-sized countries.
Alphabet's Whiplash: Praised in April, Punished in July
Google parent Alphabet's stock reaction to its own capex increases has swung dramatically within the same year, which is itself the most interesting data point here. In April 2026, Alphabet raised its capex guidance and reported Google Cloud sales up a record 63%, sending shares more than 6% higher in after-hours trading. By July 22, Alphabet raised guidance again — its third increase of 2026, from an initial $175-185 billion range up to $195-205 billion — alongside a genuinely strong quarter: revenue of $119.8 billion, up 24% year-over-year and above consensus, with Google Cloud revenue accelerating to $24.8 billion, up 82%. This time, the stock fell nearly 7%, pushing Alphabet below the $4 trillion market-cap threshold, as the company reported negative free cash flow of $5.9 billion for the quarter — its first negative free cash flow quarter since 2004. The underlying business results were arguably stronger in July than in April. The market's patience for capex increases without immediate cash-flow payoff had simply worn thinner in the intervening months.
Microsoft's Accounting Adjustment
Microsoft took a notably different path through the same earnings season. Rather than matching Alphabet's aggressive guidance increase, Microsoft held roughly steady on its 2026 capex outlook, and CFO Amy Hood disclosed an accounting change that extends the estimated useful life of Microsoft's data centers and office buildings from 15 to 25 years — a shift that moves more future data center leases from finance leases (which count toward capex) to operating leases (which don't), trimming the headline capex figure by roughly $15 billion without actually reducing real spending. Hood separately guided full calendar-2026 capex to roughly $190 billion, including about $25 billion attributable to higher component pricing (memory chips among them, connecting directly to the memory shortage covered in the first entry of this series). Microsoft's cushion against the kind of punishment Alphabet received came from a clearer revenue story: Azure grew 43% year-over-year, with trailing-twelve-month revenue surpassing $100 billion — a metered, external, rapidly growing business investors could point to directly as justification for the spending.
Meta's Problem: Spending Without a Clear Payoff Story
Meta's capital spending trajectory is the steepest of the group in percentage terms: after roughly $75.8 billion in the past year, analysts now expect approximately $135.2 billion over the next twelve months, a 79% jump, followed by another 29% increase the year after. But Meta lacks what Microsoft and Amazon have: a clean, externally metered revenue line tied directly to that AI spending. Meta's AI investment routes mostly into internal advertising-system efficiency and still-early business-AI products, neither of which gives investors a simple number to point to the way Azure or AWS revenue does. The result: Meta reported an EPS miss alongside a 91% year-over-year decline in free cash flow, and the stock was punished accordingly — not because Meta spent more than Microsoft in absolute terms, but because investors couldn't as easily see what the spending was buying.
The Real Pattern: "Spend-to-Revenue Legibility"
Across this earnings season, the direction of capex spending — up for all four companies — didn't predict how the market reacted. What did predict it was whether each company could point its AI spending at a metered, external, growing revenue line. Microsoft could, through Azure. Amazon could, through AWS, which grew 37% year-over-year at a 39.4% operating margin. Meta largely couldn't, and paid for it. Alphabet sits in an unusual middle position: its Cloud growth was legible and genuinely strong, but the market weighted the shock of a first-since-2004 negative free cash flow quarter more heavily than the underlying growth number. Apple occupies a different risk category entirely — its AI strategy runs largely through a partnership fee paid to Google rather than owned infrastructure, and its own weaker guidance (9-11% revenue growth versus roughly 12% expected) specifically cited rising memory-chip and packaging costs, the same AI-driven memory shortage covered in the first entry of this series, but from the paying-customer side of that trade rather than the beneficiary side.
2026 AI Capex and Stock Reaction at a Glance
| Company | 2026 capex trajectory | Market reaction | Why |
|---|---|---|---|
| Alphabet (Google) | Raised guidance 3 times to $195-205B | Fell ~7% in July despite a strong quarter | First negative free cash flow quarter since 2004 |
| Microsoft | ~$190B guided; accounting change trims headline by ~$15B | Held up better; stock pressure milder | Azure's 43% YoY growth gives a clear revenue story |
| Meta | ~$135.2B next 12 months, a 79% jump | EPS miss; stock punished | No clear metered external revenue tied to AI spend |
| Amazon | Elevated spending alongside AWS growth | More stable reaction | AWS grew 37% YoY at a 39.4% operating margin |
Frequently Asked Questions
How much are Google, Microsoft, Amazon, and Meta spending on AI in 2026?
Combined, the four companies plan to spend roughly $725 billion on capital expenditures in 2026, a 77% increase from 2025's $410 billion, driven almost entirely by AI data center and infrastructure buildout.
Why did Alphabet's stock fall after a strong earnings report?
Despite revenue growing 24% year-over-year and Google Cloud revenue up 82%, Alphabet reported negative free cash flow of $5.9 billion for the quarter, its first since 2004, after raising its 2026 capex guidance for the third time that year, which outweighed the strong underlying growth in investors' eyes.
Why did Microsoft avoid the stock punishment Alphabet received?
Microsoft could point to Azure's 43% year-over-year revenue growth as a clear, metered justification for its AI spending, giving investors a more direct line between capital expenditure and revenue than Meta, whose AI spending routes mostly into internal efficiency gains rather than an external revenue line.
Is this article investment advice?
No. This is a factual summary of publicly disclosed earnings results and stock price reactions, not investment advice. Stock prices are volatile; anyone considering an investment decision should consult primary financial filings and a licensed financial advisor.