Microsoft cloud revenue rises 27% as AI data center buildout continues

Microsoft cloud revenue grew 27% to $59.3 billion in its fiscal fourth quarter, executives said Wednesday, as demand for Azure and the company’s first-party AI products continued to pull through infrastructure spending. The company also added 31 data centers in the quarter, bringing its total additions this year to 88, as it builds capacity for AI workloads.

The results give Microsoft more evidence that its AI-heavy cloud spending is translating into revenue, though the company did not provide a Microsoft-only capital expenditure figure in the comments reported from the call. CFO Amy Hood said Microsoft is positioned well on infrastructure investment despite volatility in AI demand, arguing that the company’s wider product and customer base gives it room to absorb shifts if demand no longer exceeds supply.

Azure crossed $100 billion in revenue for the first time, according to Hood. She tied cloud growth to customers shifting more work to Azure, continued AI infrastructure investment and stronger usage across Azure and Microsoft 365 Commercial cloud.

How much cloud revenue did Microsoft report?

Microsoft reported $59.3 billion in cloud computing platform revenue for the quarter, up 27%. Executives said the growth came from Azure demand and Microsoft’s own AI applications and services.

Microsoft 365 Copilot, the company’s AI work assistant, now has more than 30 million paid seats, Hood said. That is up from more than 20 million paid seats in the prior quarter, giving Microsoft a larger paid base for one of its most visible enterprise AI products.

CEO Satya Nadella said Microsoft is focused on helping customers convert AI deployments into measurable business outcomes across agentic experiences, AI platforms and infrastructure. That claim tracks the broader enterprise AI shift from pilots toward production use, but Microsoft did not disclose how much Copilot usage contributed to revenue in the quarter.

Microsoft’s AI infrastructure bet stays in place

The company’s spending posture comes as large cloud and platform companies continue to raise capital expenditure plans to compete for AI workloads. Amazon, Microsoft, Meta and Google are planning up to $725 billion in capital expenditures in 2026, according to Business Insider’s reporting on recent earnings calls.

Hood said Microsoft’s backlog and newly added remaining performance obligations came from across its product and customer portfolio. Her argument is that Microsoft is less exposed to a single AI demand curve than a narrower infrastructure provider would be.

Earlier this month, Microsoft introduced Frontier Company, an engineering organization meant to place 6,000 engineering experts with customers to co-design and improve AI systems. Nadella said Microsoft had tested the model over the past year, completing more than 330 projects across 164 customers.

Nadella also said Microsoft is getting better returns from existing infrastructure by optimizing across silicon, systems and software. The company is also expanding sovereign AI offerings, a priority he said customers are raising. Sovereign AI generally refers to running AI systems under specific data residency, control or regulatory requirements, which matters most for regulated industries and governments.

Microsoft said it has seen a fivefold increase in customers building with models from multiple providers. Nadella cited Levi Strauss & Co. as an example, saying the company is using OpenAI and Anthropic models on Foundry to bring more than 1,000 domain-specific agents into one enterprise AI platform.

The company is also working with Mistral to bring its models to Microsoft Sovereign Cloud. Microsoft said that will allow customers to run the models in public, customer-controlled and disconnected environments.

This story draws on original reporting from CIO Dive.