Microsoft has reportedly scrapped leases for significant data center capacity in the U.S., sparking speculation about a potential oversupply as the tech giant expands its artificial intelligence infrastructure. Analysts at TD Cowen shared this assessment.
Investor skepticism has been mounting regarding the massive investments U.S. tech firms are making in AI infrastructure. This is partly due to slower-than-expected returns and the emergence of breakthroughs from Chinese startup DeepSeek, which has demonstrated AI technology that rivals or surpasses its Western counterparts at a significantly lower cost.
According to analysts led by Michael Elias, Microsoft, headquartered in Redmond, Washington, has cancelled leases for “a couple of hundred megawatts” of capacity with at least two private data-center operators. This information stems from supply chain checks. The analysts also noted that Microsoft has paused the conversion of statements of qualifications, which are preliminary steps towards formal leases. They pointed out that other tech companies, including Meta Platforms, have previously taken similar actions to reduce capital expenditures.
The TD Cowen note was published late on Friday and gained traction on social media over the weekend. Several media outlets covered the development earlier this week. Microsoft, the primary backer of OpenAI, the company behind ChatGPT, did not immediately respond to a request for comment. Jefferies, which is currently hosting Microsoft’s investor relations team in Sydney, stated that the company has strongly denied any changes to its data center plans.
Microsoft’s shares, which underperformed most major tech stocks last year, experienced a slight increase of roughly 0.5% in premarket U.S. trading.
These lease cancellations would represent a notable shift for a company that, just months ago, allocated $80 billion in capital expenditure for the current fiscal year, with a majority of it dedicated to AI. Microsoft has consistently emphasized that this spending is vital to overcome supply bottlenecks and effectively meet the escalating demand for AI resources.
Mark Moelder, an analyst at Bernstein, suggested that the news potentially indicates a decrease in demand, particularly after somewhat disappointing quarterly results from major cloud computing companies. However, Moelder proposed that it might also reflect the substantial capacity build-up Microsoft has undertaken in recent years. “Microsoft needed to meet demand and had a great deal of difficulty finding capacity. Management may, therefore, have rented, even at a meaningful premium, data centers and GPU capacity and negotiated more deals for additional future capacity than they needed,” Moelder stated.
