SAN FRANCISCO—Microsoft projected its fiscal year 2027 capital expenditures for AI infrastructure to reach $35 billion, a figure that gives investors new clarity on its spending strategy. The announcement, made during a special investor briefing Wednesday, sent the company's stock higher as analysts absorbed the detailed plan.
Microsoft shares rose 2.8 percent in extended trading following the briefing, adding over $150 billion to its market capitalization. The gain contrasts with months of investor anxiety over the escalating costs of AI development. Analysts have questioned whether cloud margins can hold as companies pour billions into GPUs and data centers.
Market bears had pointed to Microsoft's rising capital intensity as a drag on its Azure cloud segment. The company's prior quarter saw capital expenditures for its Intelligent Cloud unit jump 30 percent year-over-year to $14 billion, fueling speculation that heavy investments in Nvidia GPUs and custom silicon would compress operating margins for an extended period.
During the briefing, Microsoft executives outlined how the $35 billion allocation would directly support Azure AI services expansion and new Copilot features. CEO Satya Nadella said the spending targets specific revenue-generating opportunities rather than foundational research, and highlighted a strong pipeline of enterprise customers ready to adopt AI-powered solutions.
The company expects a large portion of the investment to drive higher consumption of its Azure AI platform, where customers pay for compute on a usage basis. Microsoft also anticipates increased subscription revenue from its M365 Copilot offerings, priced at $30 per user per month—a model that allows direct monetization of the infrastructure buildout.
The plan positions Microsoft to hold its ground against Amazon Web Services and Google Cloud. AWS has also committed substantial resources to AI infrastructure, reporting record capital expenditure figures in its latest earnings. Compute capacity has become the defining battleground in cloud market share.
The $35 billion budget covers advanced GPUs—primarily from Nvidia—alongside custom AI chips designed in-house and the construction and upgrade of data centers globally to handle growing AI workloads.
"Microsoft provided the granular detail investors needed to understand the path to monetization," said analyst Mark Mahaney of Evercore ISI during a client call. "The clarity around unit economics and deployment strategy changes the narrative from cost burden to strategic advantage." Mahaney maintained his Outperform rating on Microsoft stock.
Questions remain about the long-term return on capital for investments at this scale. The rapid pace of AI innovation means hardware can become obsolete quickly, requiring continuous upgrades. The ability to pass rising compute costs to enterprise customers remains the key variable in the profitability equation.
