Microsoft is canceling a significant number of data center leases in the United States, which may signal concerns about overbuilding AI computing capacity or a strategic shift toward optimizing existing resources, according to TD Cowen.
OpenAI’s largest backer has canceled data center leases totaling “a couple of hundred megawatts” of capacity, TD Cowen reported Friday, citing supply chain inquiries.
The brokerage also noted that Microsoft has halted conversions of statement of qualifications—agreements that typically lead to formal leases—a strategy previously used by Meta when reducing capital spending.
Microsoft’s pullback on spending and data center construction raises questions about whether the AI frontrunner is becoming more cautious about demand.
The company plans to spend $80 billion on AI data centers this fiscal year, and CEO Satya Nadella recently emphasized the need for sustained investment to meet “exponentially more demand.”
Microsoft reaffirmed its fiscal year spending target in a statement on Monday but declined to comment on TD Cowen’s report.
“While we may strategically pace or adjust our infrastructure in some areas, we will continue to grow strongly in all regions,” a company spokesperson said in the statement. “Our plans to spend over $80B on infrastructure this FY remains on track as we continue to grow at a record pace to meet customer demand.”
Critics continue to highlight the lack of real-world AI applications, even as Microsoft, Meta, and Amazon commit billions to AI data centers.
Wall Street’s scrutiny of these investments has intensified, especially after Chinese startup DeepSeek unveiled an open-source AI model that claims to rival U.S. technology at a significantly lower cost.
“While we have yet to get the level of color via our channel checks that we would like into why this is occurring, our initial reaction is that this is tied to Microsoft potentially being in an oversupply position,” TD Cowen analysts Michael Elias, Cooper Belanger, and Gregory Williams emphasized that their assessment was simply their interpretation.
