Microsoft’s stock declined approximately 2% on Monday morning following an analyst report suggesting the tech giant has canceled leases for several US data centers, potentially signaling softer-than-expected AI demand. According to a note from TD Cowen analysts published Friday, channel checks revealed that Microsoft has taken several significant steps to adjust its data center strategy.
The analysts reported that Microsoft has canceled leases totaling “a couple of hundred megawatts” with at least two private data center operators in the United States. Additionally, the company has pulled back on converting statements of qualifications (SOQs) to actual leases—a concerning sign since SOQs typically precede lease signings and signal green lights for data center construction. Perhaps most notably, Microsoft has reallocated a considerable portion of its international spending to the US, suggesting what analysts characterized as “a material slowdown in international leasing.”
In some instances, Microsoft reportedly used “facility/power delays as a justification for the termination” of these agreements. The TD Cowen analysts concluded that these observations collectively suggest Microsoft “may have excess data center capacity relative to its new forecast,” pointing to a potential oversupply position for the company.
In a follow-up note published Monday, the analysts offered additional context, suggesting Microsoft’s actions may be “related to a shift in incremental OpenAI workloads to Oracle/SoftBank.” They noted a material ramp in Oracle requirements over the past three months at an unprecedented scale. This development comes after President Donald Trump announced Stargate in January, a joint venture between OpenAI, Oracle, and SoftBank expected to invest up to $500 billion in AI infrastructure in the United States.
The analysts acknowledged that while Microsoft remains active in data center leasing, “their run-rate data center demand is now lower vs. the torrid pace of leasing seen in 2023 and 1H24, during which time they led the data center leasing league tables amongst hyperscalers.”
Microsoft responded to the concerns, with a spokesperson telling Business Insider that the company’s planned $80 billion spending on AI data centers this fiscal year “remains on track.” The spokesperson emphasized that Microsoft added more capacity last year than any prior year in history and is “well positioned to meet our current and increasing customer demand.” They added that while the company may “strategically pace or adjust our infrastructure in some areas,” it will “continue to grow strongly in all regions.”
Key Quotes
Our channel checks indicate that MSFT has 1) cancelled leases in the US totaling ‘a couple of hundred MWs’ with at least two private data center operators, 2) has pulled back on the conversion of SOQ’s to leases, and 3) has re-allocated a considerable portion of its international spend to the US.
TD Cowen analysts detailed their findings from supply chain channel checks, providing specific evidence of Microsoft’s data center strategy shift that triggered market concerns about AI demand.
Us self-claiming some AGI milestone, that’s just nonsensical benchmark hacking to me. The real benchmark is, is the world growing at 10%?
Microsoft CEO Satya Nadella emphasized his focus on tangible economic outcomes rather than abstract AI milestones, suggesting a pragmatic approach to measuring AI’s real-world value and impact.
You can go off the rails completely when you are hyping yourself with the supply side, versus really understanding how to translate that into real value to customers.
Nadella warned against excessive supply-side optimism in AI infrastructure investment, advocating for demand-driven growth—a statement that now appears prescient given the recent data center lease cancellations.
Thanks to the significant investments we have made up to this point, we are well positioned to meet our current and increasing customer demand. Last year alone, we added more capacity than any prior year in history.
A Microsoft spokesperson defended the company’s position, framing the lease cancellations as strategic adjustments rather than signs of weakening AI demand, while reaffirming the $80 billion spending commitment.
Our Take
This story reveals the delicate balancing act facing AI infrastructure leaders as they navigate unprecedented demand uncertainty. Microsoft’s moves appear less about retreating from AI and more about strategic recalibration—shifting workloads to partners like Oracle through Stargate while optimizing its own footprint. Nadella’s recent philosophical comments about avoiding “hype” and focusing on real customer value now seem like deliberate expectation-setting rather than casual observations. The international-to-US reallocation is particularly telling, suggesting geopolitical factors and domestic policy incentives are reshaping AI infrastructure geography. What’s most significant is the timing: if Microsoft, with its OpenAI partnership and Azure dominance, is pumping the brakes, it signals the industry is entering a more mature, efficiency-focused phase. The real test comes Wednesday with Nvidia’s earnings—if chip demand remains strong despite data center pullbacks, it suggests workload optimization rather than demand collapse.
Why This Matters
This development represents a critical inflection point for the AI infrastructure boom that has dominated tech investment over the past two years. Microsoft’s data center pullback—whether strategic repositioning or demand-driven—offers the first tangible evidence that even AI leaders are recalibrating their massive infrastructure bets. The timing is particularly significant as it coincides with the Stargate announcement, suggesting a potential reshuffling of AI workload distribution among major players rather than an overall market contraction.
For the broader AI industry, this signals a maturation phase where companies are moving from “build it and they’ll come” strategies to more measured, demand-aligned approaches. CEO Satya Nadella’s recent comments about supply-demand mapping and avoiding “hype” underscore this shift toward sustainable growth. The reallocation from international to US markets also reflects geopolitical considerations and the impact of initiatives like Stargate on AI infrastructure geography. Wall Street will be watching Nvidia’s Wednesday earnings closely for corroborating signals about AI spending trends, making this a pivotal week for assessing the health and trajectory of AI infrastructure investment.
Recommended Reading
For those interested in learning more about artificial intelligence, machine learning, and effective AI communication, here are some excellent resources: