Microsoft has been pulling back on data center projects around the world, prompting questions about the company’s expectations for future demand of cloud and AI services.
Bloomberg reports Microsoft, widely seen as a leader in commercializing AI services, has recently halted or delayed development of data center sites in several countries, including Indonesia, the UK, Australia, and the United States. This pullback has raised concerns among investors who closely track the company’s spending plans to gauge long-term customer demand for cloud and AI services.
While it is unclear whether the retrenchment reflects diminished demand expectations or temporary construction challenges, such as power and building material shortages, some investors have interpreted the signs as an indication that projected AI service purchases may not justify Microsoft’s massive server farm outlays. These concerns have weighed on global tech stocks in recent weeks, particularly chipmakers like Nvidia, which account for a significant share of data center budgets.
Microsoft acknowledged making changes to its data center plans but declined to discuss most of the projects in detail. A spokesperson stated, “We plan our data center capacity needs years in advance to ensure we have sufficient infrastructure in the right places. As AI demand continues to grow, and our data center presence continues to expand, the changes we have made demonstrate the flexibility of our strategy.”
In the UK, Microsoft withdrew from negotiations to lease space at a site marketed for its ability to host advanced Nvidia chips. The company also halted negotiations for data center space near Chicago and backed away from a proposal to obtain more capacity from CoreWeave, a company from which it had previously leased excess cloud-computing capacity.
In some cases, Microsoft is delaying construction, such as pausing work on parts of a data center campus outside Jakarta, Indonesia. The company has also put on hold some planned expansion at a site in Mount Pleasant, Wisconsin, part of a complex visited by then-President Joe Biden.
Analysts have stepped up their scrutiny of data center spending since Chinese upstart DeepSeek announced in January that it had created a competitive AI service using fewer resources than leading US companies. This development suggests that new engineering techniques could mean AI will require less computing power than previously expected in the long run.
TD Cowen analysts wrote last week that Microsoft has abandoned new data center projects in the US and Europe amounting to a capacity of about two gigawatts of electricity, likely representing “data center oversupply relative to its current demand forecast.” The analysts also noted that the pullback might reflect Microsoft’s choice to forgo some new business from OpenAI, the leading AI startup now valued at $300 billion, which recently unveiled a joint venture with SoftBank Group and Oracle to invest between $100 billion and $500 billion in AI infrastructure.
Alibaba Chairman Joe Tsai warned of a potential bubble forming in data center construction in late March, arguing that the pace of the buildout may outstrip initial demand for AI services. CoreWeave’s CEO, Michael Intrator, said that the pullback in data center spending is more specific to Microsoft than the larger industry, citing changes in the company’s relationship with OpenAI.
Despite the pullback, Microsoft remains committed to spending about $80 billion on building out data centers in its current fiscal year ending in June. However, the company has previously stated that the following fiscal year will see a slower rate of new infrastructure spending, with a focus on fitting out existing facilities with servers and other equipment rather than new construction.
Read more at Bloomberg here.
Lucas Nolan is a reporter for Breitbart News covering issues of free speech and online censorship.