Microsoft posted record fiscal 2026 revenue as Azure crossed $100 billion and proved its AI bet is paying off

Microsoft closed fiscal 2026 with $331.8 billion in revenue, up 18%, as Azure crossed $100 billion in annual revenue for the first time with 41% growth, accelerating to 43% in the June quarter, while Microsoft 365 Copilot passed 30 million paid seats, up from more than 20 million three months earlier. The results provided investors with proof that massive AI spending can generate measurable revenue, contrasting with Meta's same-day report of a 14% net income drop despite 28% revenue growth and a raised capex outlook of $130-$145 billion for 2026.

Microsoft's fiscal 2026 results gave investors what Meta still hasn't: proof that massive AI spending can already sit on top of a working revenue machine. The numbers are hard to argue with. Microsoft closed the fiscal year ended June 30, 2026 with $331.8 billion in revenue, up 18%, according to the company's results cited by the Associated Press. Azure crossed $100 billion in annual revenue for the first time, grew 41% across the year, and accelerated to 43% in the June quarter. Microsoft 365 Copilot passed 30 million paid seats, up from more than 20 million three months earlier. That's the part you should watch. It isn't a pitch deck. It's paid adoption. On the same evening, Meta gave the market a very different kind of AI story. Axios reported that Meta's expenses jumped 55% to about $42 billion, while net income fell 14% to $15.8 billion even as revenue rose 28%. MarketWatch reported that Meta lifted its 2026 capital expenditure outlook to between $130 billion and $145 billion. The stock fell after the report. Meta is spending as if the return is obvious, but its consumer AI products still don't have the same direct paid lane that Microsoft gets from Azure, Microsoft 365 and enterprise security. Microsoft has that lane. Meta doesn't. The capex question Azure's 43% June-quarter growth is the cleanest answer Microsoft could have given to the AI spending sceptics. Large companies aren't merely testing cloud AI tools in a corner of the IT budget. They're buying more capacity and more software seats, and Microsoft is turning that demand into a business line Wall Street can measure. AP reported that Microsoft guided for Azure growth of about 45% in constant currency for the September quarter, ahead of the 43% just posted. That is acceleration on a very large base. The spending is still enormous. Axios reported that Microsoft recorded $41 billion in capital expenditures in the quarter, up 70%, with about two-thirds going to short-lived assets such as CPUs and GPUs. Earlier in fiscal 2026, Microsoft CFO Amy Hood told investors the company expected roughly $190 billion of calendar-year 2026 capital expenditures before an accounting adjustment brought the effective forecast to about $175 billion, according to Business Insider. That adjustment matters, but don't overread it. Microsoft isn't suddenly building a small AI business. It is still building one of the largest infrastructure programs in corporate history. Investors have been asking the right question for months: when does the spending pay back? In January, Fortune quoted Morgan Stanley analyst Keith Weiss telling Microsoft's executives that capex was growing faster than expected while Azure was growing a little slower than expected, raising concern about return on investment. That was a fair worry then. It is a weaker worry after this print. Hood's answer has been consistent in Microsoft's own earnings calls. She has pointed to demand exceeding supply, Microsoft 365 Copilot growth, GitHub Copilot, first-party AI products and long-term research needs as reasons the company doesn't allocate every new GPU directly to Azure. You may not love the scale of the bill, but at least you can see the machinery that is meant to repay it. There are also small, concrete signs of usefulness beneath the headline figures. Microsoft's own Security Copilot adoption materials say its phishing alert triage agent helped users triage submitted phishing alerts 78% faster, with more accurate verdicts. Accenture's Copilot rollout is larger still: Microsoft said in April that Accenture was rolling Copilot out to around 743,000 workers, its largest enterprise Copilot deployment at the time. Those examples aren't the whole market, but they show why a finance chief can defend the spend without sounding like she is selling a theory. Enterprise adoption is the difference The real distinction between Microsoft and Meta is not that one spends and the other doesn't. Both spend heavily. The distinction is where the spending lands. Microsoft can sell AI into companies that already pay for Microsoft 365, Azure, Entra, Teams, Outlook, GitHub, Dynamics and security software. The customer relationship already exists. The bill already exists. AI becomes another line on it. Meta's position is structurally harder. Its AI tools sit inside consumer products where users don't usually pay a subscription fee, and its infrastructure spending has to work its way back through advertising, engagement, agent commerce or some future compute business that still needs proving. Frankly, that is a tougher bridge. Meta may eventually find a huge business in personal AI agents, but Microsoft's fiscal 2026 results show revenue converting now, not in a future strategic narrative. Copilot's 30 million paid seats help make that point. TechCrunch reported after Microsoft's March-quarter call that Satya Nadella said Copilot engagement was rising and that users were engaging with it about as much as email. By the June quarter, the paid-seat count had climbed again. You don't need to pretend every licensed employee is transformed by it. Some won't be. But when paid seats rise by roughly 10 million in a quarter, the market has to treat enterprise AI software as a real revenue product, not only a demo. The $331.8 billion revenue figure is the headline. Azure crossing $100 billion is the story. A cloud business that large, still growing above 40%, gives Microsoft a cushion Meta doesn't have and a clearer route from data-center spending to customer revenue. The capex debate isn't over. It shouldn't be. But after these results, Microsoft has earned a stronger answer than most of its AI peers: the spending is already attached to customers who pay. Also read: IonQ wins FTC clearance to close its $1.8 billion SkyWater acquisition and become America's only quantum chipmaker https://startupfortune.com/ionq-wins-ftc-clearance-to-close-its-18-billion-skywater-acquisition-and-become-americas-only-quantum-chipmaker/ • Zuckerberg bets billions of personal AI agents on Meta's most expensive quarter ever https://startupfortune.com/zuckerberg-bets-billions-of-personal-ai-agents-on-metas-most-expensive-quarter-ever/ • Manhattan Associates stock surged 27% on July 29 after a blowout Q2 2026 earnings report https://startupfortune.com/manhattan-associates-stock-surged-27-on-july-29-after-a-blowout-q2-2026-earnings-report/