Microsoft has reportedly cancelled leases for a significant data centre capacity in the US, raising concerns about the long-term sustainability of AI infrastructure investments, according to Bloomberg, which cites analyst TD Cowen.
The report added that the tech giant cancelled agreements of “a couple of hundred megawatts” of capacity, citing insights from supply chain providers.
The decision suggests Microsoft may be reassessing its AI computing needs, even as it pledges to spend $80 billion this fiscal year on expanding computing capacity.
The report, released on Friday, noted that Microsoft has also halted conversions of statement of qualifications, a step that typically leads to formal leasing agreements. The move has sparked speculation over whether Microsoft is adjusting its AI strategy due to potential overcapacity.
On its January earnings call, CEO Satya Nadella said that the company must sustain high levels of spending to meet “exponentially more demand”. However, Wall Street has increasingly questioned the long-term viability of such investments, given the uncertain commercial applications of AI.
In response to the report, Microsoft reiterated its spending target but acknowledged some adjustments in infrastructure development.
“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. “Our plans to spend over $80 billion on infrastructure this fiscal year remain on track as we continue to grow at a record pace to meet customer demand.”
At the same time, OpenAI appears to be exploring alternative computing options. In a recent report, The Information suggested that OpenAI is planning to shift its workload from Microsoft to Project Stargate. The report states that in recent weeks, OpenAI has informed investors that Stargate—a developing data centre expansion initiative, expected to receive substantial funding from SoftBank—could supply around 75% of the computing power needed to operate and refine its AI models by 2030.
Notably, Microsoft is no longer OpenAI’s exclusive cloud partner. In a recent blog, OpenAI announced a new large-scale commitment to Azure, which will continue supporting all its products and model training. However, the agreement now allows for more flexibility.
Instead of exclusivity, Microsoft has a right of first refusal on any new capacity OpenAI wants to add. This means Microsoft gets the first chance to match any other cloud provider’s offer before OpenAI can move forward with them.
On the other hand, the emergence of cost-efficient AI models, such as the open-source model developed by Chinese company DeepSeek, has intensified scrutiny of major firms’ AI expenditures.
DeepSeek claims its model rivals US technology at a fraction of the cost, raising questions about the financial sustainability of large-scale AI infrastructure investments.
