
Microsoft’s OpenAI business reaches $10 billion a quarter
Microsoft booked $10 billion in revenue from its commercial arrangements with OpenAI in the three months ended September 30, 2026, according to the company’s quarterly filing. The figure amounts to roughly 11% of Microsoft’s total revenue for the period and is the clearest public measure yet of how much money flows through one of the technology industry’s most closely watched partnerships.
The disclosure came as Microsoft reported quarterly results and gave investors a fresh look at both the benefits and the risks of its AI strategy. OpenAI is not only a major Microsoft investment; it is also a customer of Microsoft’s Azure cloud business and a counterparty in a revenue-sharing arrangement. That dual role makes the relationship unusually difficult to evaluate, even for analysts who follow the company closely.
Microsoft had not previously broken out a quarterly revenue number tied specifically to OpenAI. The new filing therefore gives the market a sharper, if still incomplete, view of the financial ties between the two companies as they loosen some of the exclusivity that once defined their alliance.
Key facts from the filing
- Microsoft recorded $10.0 billion of revenue from commercial arrangements with OpenAI in the quarter ended September 30, 2026, including revenue-sharing payments.
- The OpenAI-related revenue was about 11% of Microsoft’s total revenue for the quarter.
- Microsoft did not split the $10 billion between revenue sharing and OpenAI’s spending on Azure.
- In the 12 months to June, Microsoft recorded $24.1 billion from OpenAI, or roughly $6 billion a quarter on average.
- The latest quarterly figure exceeded Xbox revenue of $5.4 billion and search and advertising revenue of $6.6 billion.
- OpenAI owed Microsoft $7.2 billion at the end of September, up from $6.0 billion in June.
- Microsoft owns about 24% of OpenAI and has paid $12.0 billion of the $13.0 billion it committed to invest.
Why the disclosure matters
The $10 billion figure is significant because it turns a broad narrative about Microsoft’s AI leadership into a concrete revenue number. For more than a year, investors have debated whether the massive sums being spent on data centers, chips and AI infrastructure would translate into meaningful sales. The OpenAI disclosure offers one answer: the relationship is already producing revenue at a scale comparable to some of Microsoft’s established consumer and advertising businesses.
At the same time, the filing shows how concentrated that revenue may be. Microsoft has not said how much of the $10 billion came from revenue sharing and how much came from OpenAI’s purchases of Azure computing capacity. That distinction matters. Revenue sharing is tied to OpenAI’s own sales performance, while Azure spending reflects OpenAI’s demand for cloud infrastructure. A shift in either area could change the financial picture quickly.
Under accounting rules, OpenAI counts as a related party for Microsoft, which is why the company must disclose certain financial details about the relationship. The filing shows that the connection is substantial in both directions. Microsoft earns money from OpenAI through revenue sharing and cloud services, while OpenAI owes Microsoft billions of dollars and relies heavily on its infrastructure.
A partnership that is becoming less exclusive
Microsoft’s relationship with OpenAI dates to 2019, when the software company made an initial $1 billion investment and became the AI developer’s preferred cloud partner. The partnership deepened in 2023 with a multibillion-dollar extension that made Microsoft the most prominent corporate backer of OpenAI’s large language models. Over time, the arrangement helped Microsoft move quickly into generative AI, embedding OpenAI models into products such as Azure OpenAI Service and Copilot, while giving OpenAI access to vast computing resources it needed to train and run increasingly large models.
That history helps explain why the $10 billion quarterly figure is so closely watched. Microsoft’s market valuation has been shaped by investor enthusiasm for AI, but investors have also grown impatient for concrete evidence that the spending will produce durable revenue. The OpenAI disclosure offers one such data point, although it also underscores how concentrated part of Microsoft’s AI revenue may be.
The two companies have gradually loosened the ties that once made Microsoft OpenAI’s default cloud provider. In October 2025, Microsoft gave up its right of first refusal to supply OpenAI’s computing capacity. At the same time, OpenAI committed to buy another $250 billion of Azure services, a huge signal that the relationship would remain commercially important even as it became less exclusive.
An April 2026 agreement gave OpenAI broader freedom to use other cloud providers. OpenAI has since brought some of its models to a rival cloud platform, a move that would have been harder to imagine during the early years of the alliance. The changes reflect a maturing relationship: Microsoft remains a central partner, but OpenAI is seeking more flexibility and access to computing capacity from multiple suppliers.
OpenAI will keep sharing revenue with Microsoft through 2030, subject to a cap. Earlier reports have said OpenAI shares 20% of its revenue, though neither company has confirmed that figure. The lack of a precise breakdown in the filing leaves investors to infer how much of the $10 billion came from revenue sharing and how much came from OpenAI’s payments for Azure computing.
The stake structure has also shifted. Microsoft owns about 24% of OpenAI, down from about 25% in June. After OpenAI’s restructuring in October 2025, Microsoft’s stake was about 27%. The company did not say what caused the latest dilution. Even so, Microsoft has already paid $12 billion of the $13 billion it committed to invest, making it one of OpenAI’s most important financial backers.
Nadella stresses diversification
On the earnings call, Bernstein analyst Mark Moerdler asked where Azure’s AI growth was coming from. Chief executive Satya Nadella did not mention OpenAI in his response. Instead, he said he did not want Azure’s AI business to depend on a few large customers.
Relying on “four customers who buy a lot of capacity from you” for a few quarters of growth is “not a business as far as I’m concerned,” Nadella said. The comment was notable because OpenAI is widely believed to be one of Azure’s largest AI customers. Nadella’s answer suggested that Microsoft wants to show that AI demand is spreading across a broader base of enterprise and smaller business customers, rather than being concentrated among a handful of model developers.
In a social media post after the call, Nadella said Microsoft’s infrastructure is a “fungible fleet” that allows it to dynamically allocate capacity to the highest-value opportunities across compute, data, Foundry and Copilot. He said the company is bringing capacity online faster and getting more from the infrastructure it already has. Azure growth, he wrote, was broad-based across enterprise and small and medium-sized businesses, spanning both core cloud services and AI infrastructure. He added that he had never been more confident in the opportunity ahead.
Warnings about reliance on outside models
Microsoft also widened a warning in the filing about its reliance on outside AI model makers. The company said it has “limited visibility into or control over” how those models are designed, trained, secured and tested. It added that some of those partners are also significant Azure customers, creating a complicated set of commercial and operational dependencies.
The warning is a reminder that Microsoft’s AI business is not built solely on its own technology. The company has invested heavily in OpenAI and uses OpenAI models in its products, but it also works with other model providers and develops its own models. That multi-partner approach gives Microsoft flexibility, but it also introduces risks that are harder to manage than those in a purely in-house software business.
For investors, the concern is not that the OpenAI relationship is unimportant. It is that the relationship is so important that any change in OpenAI’s strategy, financial condition or cloud preferences could affect Microsoft’s revenue and costs. The filing’s disclosure of OpenAI’s $7.2 billion debt to Microsoft highlights how intertwined the two companies remain, even as they give each other more room to work with competitors.
Capital spending forecast trimmed
Chief financial officer Amy Hood cut Microsoft’s 2026 capital spending forecast to $170 billion from $175 billion. She said some “GPU-as-a-Service” contracts with outside capacity providers had slipped to the second half of the fiscal year. The revision does not necessarily signal a pullback in Microsoft’s AI ambitions, but it does show how quickly spending plans can shift when third-party capacity arrangements change.
Capital spending, including finance leases, was $45.2 billion in the quarter, below Microsoft’s own $50 billion projection. The cash portion, spending on property and equipment, was $39.4 billion. Hood said spending would top $50 billion in the current quarter. Investor relations chief Jonathan Neilson attributed the shortfall to delays rather than any change in spending plans.
The numbers illustrate the scale of the infrastructure race in which Microsoft is engaged. Building and equipping data centers for AI requires enormous amounts of capital, and the returns depend on filling that capacity with paying customers. Microsoft’s ability to generate $10 billion a quarter from OpenAI helps offset some of that cost, but it does not eliminate the risk that AI demand may not grow as quickly as the infrastructure buildout.
Microsoft’s headcount fell 2% in the quarter. The company also offered buyouts to 60,000 employees this week, a move that suggests it is trying to control costs while continuing to spend heavily on AI infrastructure. The combination of huge capital outlays, a slightly smaller workforce and a growing but concentrated AI revenue stream is the central tension in Microsoft’s current financial story.
Source:TNW | Openai News
