Microsoft's Azure Cloud Crosses $100 Billion as AI Spending Boom Extends

Today’s Focus

Microsoft reported fiscal fourth-quarter earnings on Wednesday that pushed its annual revenue past $331 billion, with cloud unit Azure crossing $100 billion in yearly sales for the first time, according to The Wall Street Journal and Bloomberg.

Quarterly profit rose 31% from a year earlier, the Journal reported, and Azure’s growth accelerated to its fastest pace since 2022, per Bloomberg’s reporting on the results. Shares jumped roughly 8% after the release, CNBC said, briefly pushing Microsoft’s market value higher in extended trading.

Chief Executive Satya Nadella and Chief Financial Officer Amy Hood told analysts on the earnings call that capital expenditures would rise again in fiscal 2027 to meet what Hood described as demand outstripping supply for AI computing capacity, according to the Yahoo Finance transcript of the call.

Microsoft has been the largest single customer of Nvidia’s AI chips and the biggest financial backer of OpenAI, and its data center buildout has become a proxy for the health of the broader AI trade. Morningstar analyst Dan Romanoff wrote after the release that Azure’s reacceleration and the widening backlog of contracted cloud commitments suggested demand was not slowing.

CNBC reported that Microsoft’s commercial remaining performance obligations, essentially signed but not yet delivered contracts, climbed to a record level, which executives cited as justification for the higher spending plan. The company did not disclose a specific dollar figure for fiscal 2027 capex on the call but indicated the growth rate would exceed fiscal 2026, per Bloomberg.

The results landed as the Federal Reserve held interest rates steady the same day, a decision The Washington Post said reflected continued concern about inflation even as tech valuations kept climbing.

The Debate

Supporters argue

Nadella told analysts that AI workloads are producing real revenue, not just usage, and that Microsoft is “capacity constrained” heading into fiscal 2027, according to the Yahoo Finance transcript. He pointed to Azure OpenAI, GitHub Copilot, and Microsoft 365 Copilot as products with paying enterprise customers at scale.

Wedbush analyst Dan Ives, quoted by CNBC, called the quarter a validation of the multi-year AI capex cycle and said the results should quiet investors worried about spending discipline. Morningstar’s Romanoff raised his fair-value estimate on the stock, arguing that the widening contract backlog gives Microsoft visibility into revenue that competitors cannot match.

Hood said on the call that gross margins in the cloud business held up despite the depreciation weight of new data centers, which supporters read as evidence that AI is accretive rather than dilutive. Bloomberg noted that Azure’s growth rate topped Wall Street’s consensus by several percentage points, the fourth straight quarter of upside surprises.

Critics argue

Bloomberg reported that even bullish investors questioned how long Microsoft can raise capex without eventually pressuring free cash flow, particularly as the company signals another step-up in fiscal 2027. Morningstar flagged that depreciation costs from the buildout will keep climbing and could compress margins if AI revenue growth decelerates.

Short-seller Jim Chanos, quoted in a Financial Times column cited by CNBC, has argued that hyperscaler AI spending resembles the late-1990s telecom fiber glut, with capacity being built ahead of durable end-user demand. Some analysts noted on the call that Microsoft still does not break out Azure OpenAI revenue separately, making it difficult to verify how much of Azure’s growth is core cloud versus AI workloads tied to a single partner.

D.A. Davidson analyst Gil Luria told CNBC that Microsoft’s reliance on OpenAI, both as a customer and as a technology partner, concentrates risk in a company that continues to burn cash. Critics also pointed to the Fed’s decision the same day to hold rates as a reminder that higher-for-longer borrowing costs make aggressive capex bets more expensive to finance.

What the experts say

Independent research suggests the AI infrastructure buildout is unusually large by historical standards. Bain & Company’s 2025 Global Technology Report estimated that global AI-related capital spending by the largest cloud providers would exceed $500 billion cumulatively by 2027, a pace Bain analysts described as without a clear postwar precedent in enterprise IT.

The International Energy Agency projected in its April 2025 report that global data center electricity demand would roughly double by 2030, driven primarily by AI workloads, raising questions about the power grid constraints that Microsoft and its peers face.

Erik Brynjolfsson of the Stanford Digital Economy Lab has published work arguing that generative AI is producing measurable productivity gains in specific tasks such as customer service and coding, though he cautioned in a 2024 NBER paper that economy-wide productivity effects typically take years to materialize. The Congressional Budget Office noted in a January 2025 outlook that business investment in information processing equipment reached its highest share of GDP on record, which economists at the Federal Reserve Bank of Dallas have said is now a meaningful driver of overall U.S. growth.

By the Numbers

$331 billion: Microsoft’s fiscal 2026 revenue, a record, per The Wall Street Journal.

$100 billion: Azure’s annualized revenue run rate, crossed for the first time this quarter, per Bloomberg.

31%: year-over-year growth in quarterly profit, according to The Wall Street Journal.

8%: approximate share-price gain in trading after the release, per CNBC.

$500 billion: Bain & Company’s estimate of cumulative hyperscaler AI capex through 2027.

2x: projected increase in global data center electricity demand by 2030, per the International Energy Agency.

Fastest since 2022: Azure’s growth rate this quarter, per Bloomberg.