Bloomberg and CNBC reported that Microsoft is reviewing its specialist AI cloud relationships as the company disclosed $45.2 billion in quarterly capital spending and delays to some expected third-party capacity deliveries.
Bloomberg published its account Nov. 5, and CNBC followed Nov. 6, citing people familiar with the discussions rather than a public Microsoft announcement. CNBC reported Microsoft had not decided which providers to stop using. It named CoreWeave, Nebius, Chronoscale, IREN, Lambda and Nscale among Microsoft’s neocloud relationships. A Microsoft spokesperson told CNBC the company uses its own data centers, leased facilities and third-party partners; the statement did not confirm the reported review.
Microsoft Chief Financial Officer Amy Hood said capital expenditures totaled $45.2 billion in the three months ended Sept. 30, 2026, and some third-party capacity deliveries expected during the quarter were delayed. Microsoft said roughly three-quarters of its capital spending went to short-lived assets, mainly central processing units and graphics processing units.
IREN reported $200 million in AI Cloud revenue for the quarter ended Sept. 30, up 182% from the previous quarter. The company moved its more than $4 billion annualized run-rate revenue target from year-end 2026 to January 2027, with a possible extension to the end of February. IREN said the target remained contracted and funded, and that a customer’s design revision and supplier scheduling affected the timing. It did not identify the customer.
CNBC separately reported Microsoft had requested a networking-equipment upgrade for IREN capacity, shifting the expected service start for some infrastructure to January or February. IREN’s release did not identify Microsoft as the customer. CNBC also reported that neocloud providers face pressure from borrowing costs, high debt loads and uncertainty about the pace of AI infrastructure expansion. CoreWeave and Nebius were scheduled to report quarterly results the following week, CNBC reported.
