News

Microsoft's AI Datacenter Cancellations Raise Concerns Over Generative AI Growth

• By John K. Waters
• February 26, 2025

Microsoft's recent decision to cancel several U.S.-based AI datacenter leases has sparked concerns among analysts about the long-term trajectory of generative AI.

Investment firm TD Cowen revealed on Sunday that Microsoft had halted contracts with two private datacenters, amounting to hundreds of megawatts of capacity, due to power delay issues—a move the firm likened to Meta's withdrawal from datacenter leases during the scaling down of its metaverse initiatives.

"Our recent channel checks indicate that Microsoft has terminated select leases with at least two private data center operators across multiple US markets to the tune of 'a couple of hundred MW,'" wrote analysts led by Michael Elias. They also said that some of its real estate management tactics mirror measures that Meta took while curtailing its metaverse-related capex binge. Separately, they suggested that Microsoft may be reallocating some of its international spending budget to the US.

“Our initial reaction is that this is tied to Microsoft potentially being in an oversupply position,” they concluded.

The investment firm also reported that Microsoft has paused or delayed statement of qualifications (SOQ) for multiple U.S.-based datacenters, suggesting that the company may not require additional AI computing power. Although an SOQ signals intent to lease, it does not constitute a binding agreement.

Signs of AI Overcapacity?
Though TD Cowen acknowledged it lacks insider information, the firm suggested that Microsoft's actions indicate it is in an "oversupply position" on AI datacenter power. The company, which was the most aggressive lessee of capacity in 2023 and early 2024, may now be adjusting its projections based on OpenAI workloads and broader AI demand.

The report pointed to Microsoft's halted construction of a datacenter in Wisconsin, which was reportedly meant to support OpenAI's operations. This decision, TD Cowen suggested, could be evidence that Microsoft has more datacenter capacity than necessary in regions where the infrastructure cannot be repurposed for general cloud computing.

AI Growth Under Scrutiny
Further adding to concerns, Microsoft CEO Satya Nadella raised questions about AI's real economic impact in a podcast interview on Sunday. Speaking on the Dwarkesh Podcast, Nadella downplayed internal growth projections, arguing that the true measure of AI's success lies in its effect on global economic expansion.

"In 2025, as we sit here, I'm not an economist, but at least I look at it and say we have a real growth challenge," Nadella said. "If we claim this is like the Industrial Revolution, then we need to see Industrial Revolution-level growth."

According to Nadella, for AI to be considered an economic breakthrough, global GDP would need to rise by 10% due to AI-driven efficiencies—adjusted to 5% accounting for inflation. However, the current economic impact of generative AI remains far below this threshold, despite widespread adoption.

Microsoft issued a statement to Seeking Alpha clarifying the TD Cowen note:

"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. While we may strategically pace or adjust our infrastructure in some areas, we will continue to grow strongly in all regions. This allows us to invest and allocate resources to growth areas for our future. 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."

What It Means for the AI Market
Microsoft's moves could signal a cooling period for AI infrastructure expansion, following the industry's explosive growth in recent years. If AI-driven revenues fail to meet expectations, companies may begin scaling back their multi-billion-dollar investments in AI infrastructure.

Although Microsoft remains a leader in AI and cloud computing, analysts will be watching closely to see whether the company's shift in datacenter strategy is a sign of a broader industry slowdown—or merely a strategic adjustment to a rapidly evolving market.