OpenAI ignites Microsoft (MSFT.US) AI revenue super engine! The bull market in AI is shifting from “buying chips” to “monetizing AI applications”

According to Odaily, information disclosed by US technology giant Microsoft (MSFT.US) shows that most of this AI cloud infrastructure and software-focused giant’s AI-related revenue comes from OpenAI. OpenAI is not only Microsoft’s most important cutting-edge AI large model provider but also a super client, consuming Azure computing power, paying for model training and massive AI inference fees, and sharing revenue and profits; as a major shareholder, Microsoft utilizes Azure infrastructure, enterprise sales channels, and software entry points such as Microsoft 365, GitHub, and Dynamics, to package OpenAI models into scalable enterprise AI application service ecosystems.

It is reported that Microsoft recently stated in a regulatory filing that in the fiscal year ending in June, the tech giant obtained about $24.1 billion in revenue from this AI startup. Microsoft CEO Satya Nadella had previously said that by the end of the fiscal quarter in March, the company’s AI business, at its then growth rate, was on pace to hit $37 billion in annual revenue. In last week’s release of fourth-quarter results, Microsoft did not update total sales figures for its AI-related business.

The main investment narrative in the stock market is now gradually shifting from “who builds the largest GPU data center” to “who can convert Tokens into sustainable cash flow.” This is the fundamental logic behind Goldman Sachs’ belief that Microsoft is the core beneficiary in the “AI application monetization stage.” Compared with reacquiring customers for independent AI applications, Microsoft can install Copilot and agent services for its massive existing enterprise base through cross-selling, achieving lower marginal customer acquisition costs, deeper data and workflow moats, and at the same time, covering models, cloud platforms, developer tools, and a super-ecosystem of AI applications.

The Truth Behind the AI Revenue Boom: OpenAI May Account for About 70% of Microsoft’s AI Revenue

This latest disclosure indicates that in AI-related business revenue actually realized in Microsoft’s most recent fiscal year, OpenAI contributed more than half and likely around 70%, highlighting the extent to which Microsoft remains dependent on this close partner.

According to the agreement between the two companies, OpenAI is required to pay Microsoft for AI cloud computing infrastructure, for costs incurred in building AI models, and to share a proportion of system revenues. Microsoft has been trying to reduce its reliance on this partner, including investing in OpenAI rival Anthropic and developing its own models. However, when assessing Microsoft’s AI business progress and overall valuation, investors keep asking: exactly how much of this revenue is contributed by OpenAI?

Assumptions by market research firm Bloomberg Intelligence indicate that at the annual run rate (ARR) of Microsoft’s AI business announced in March—123% growth—Microsoft’s AI-related business would have generated around $34 billion in the fiscal year ending June. This estimate allows a direct comparison with OpenAI’s approximately $24.1 billion contribution to Microsoft that fiscal year as newly disclosed.

Microsoft has disclosed its total AI-related business size only twice. The first time was for the quarter ending December 2024, when the company announced that its AI-related business was on pace to exceed $13 billion in annual sales. The second announcement was in the earnings call for the March quarter, when Microsoft said AI-related business revenue for the year could exceed $37 billion.

The total amount for Microsoft’s AI-related business broadly includes revenues from all AI-related client groups and sales of Microsoft’s AI-specific products to any customer. However, a Microsoft spokesperson previously confirmed that the OpenAI revenue contribution figure primarily covers all sales and revenue sharing from OpenAI.

When compared with Microsoft’s total revenue, OpenAI’s share appears much smaller, less than 10%. Microsoft announced that commercial orders added in the latest quarter totaled approximately $51 billion, mainly driven by customers other than AI startups. Nevertheless, OpenAI still contributed most of Microsoft’s annual order growth.

Until last week, Microsoft had never explicitly disclosed all of its revenue received from OpenAI. Olga Usvyatsky, accounting researcher and founder of data analytics firm Nonlinear Analytics, wrote in a research report that this latest disclosure may be related to OpenAI’s preparations for its initial public offering (IPO).

Jackson Ader, a senior analyst at Wall Street financial giant KeyBanc, stated a major question remains unanswered: in the OpenAI revenue contribution, how much comes from revenue sharing agreements, and how much is from AI cloud services or other offerings by Microsoft. He commented: “The greater the portion of this revenue that comes from providing services to OpenAI, rather than investment gains, the more positively I view this business.”

From "Selling AI Compute Power" to "Selling Productivity": OpenAI Fuels Microsoft’s AI Monetization Engine

Microsoft’s heavy reliance on OpenAI is not due to a lack of cloud computing or software engineering capabilities, but because the two parties jointly built a highly-coupled “bilateral business flywheel”: OpenAI is not only Microsoft’s most important frontier model provider but also a super client that consumes Azure computing power, pays model training and inference bills, and shares revenue; meanwhile, Microsoft uses Azure infrastructure, enterprise sales channels, and software entry points like Microsoft 365, GitHub, and Dynamics to package OpenAI models into scalable enterprise services.

The latest disclosures show that in the year ending June 2026, Microsoft recognized approximately $24.1 billion in revenue from commercial arrangements related to OpenAI; according to Bloomberg Intelligence estimates, this may account for about 70% of Microsoft’s actual AI sales. Thus, Microsoft’s current AI growth benefits from OpenAI’s technological leadership, but also faces risks of client concentration, bargaining power, and revenue cyclicality. Investors must distinguish between compute and revenue sharing from OpenAI itself and true AI application terminal revenues from millions of independent enterprise clients.

Microsoft’s latest quarterly revenue hit $90 billion, up 18% year-on-year; Azure and other cloud services grew 43%, and annual revenue from Azure exceeded $100 billion for the first time. Microsoft 365 Copilot’s paid seats rose from over 20 million in the prior quarter to more than 30 million. At the AI application layer, Microsoft’s core advantage is not just models but a firm grip on enterprise identity permissions, email, documents, meetings, code, CRM, and data governance systems, enabling direct AI integration into existing workflows, and through per-seat subscriptions, usage billing, and Azure consumption, forming a highly sticky recurring revenue stream.

The AI-driven super bull market is shifting from “buying chip stocks” to “buying AI workflows,” refocusing investing themes from “who spends the most on CapEx” to “who can most rapidly convert compute into ARR, margins, and free cash flow.” This latest rotation benefits software companies that are embedded in key enterprise processes, have high renewal rates, data moats, and monetizable agents focused on AI applications, though it does not guarantee that all traditional software stocks will rise together.

As Microsoft, Amazon, and Google see their stock prices soar—especially Amazon’s 20% surge since late July, pushing its market cap above $3 trillion—software stocks have recently outperformed both the broad market and semiconductors. In July, the S&P 500 edged down about 0.1%, while the iShares Software ETF (IGV) rose 4.4%; Workday, Accenture, and Cognizant were up about 31%, 33%, and 43% respectively for the month. On some trading days marked by strong deleveraging and chip stock selloffs, application software names like Adobe, ServiceNow, Workday, and Palantir rose 7–10% in a single day, with the software ETF up 3.3% while the semiconductor ETF fell.

Goldman Sachs’s logic for favoring Microsoft as the core beneficiary of the "application monetization stage" is that, compared to reacquiring customers for independent AI solutions, Microsoft can install Copilot and agent services alongside its massive existing enterprise base at lower marginal acquisition costs, deeper data and workflow moats, and cover everything from models and cloud platforms to development tools and applications. But investment decisions can't be based solely on AI business top-line revenue; whether Microsoft can deliver another round of valuation re-rating depends on AI revenue growth after excluding OpenAI revenue sharing, Copilot paid seats and per-seat revenue, enterprise renewal rates, Azure AI gross margin, and return on capital expenditures. In other words, OpenAI got Microsoft a ticket to the AI era, but whether Microsoft is the long-term biggest winner depends on its ability to transform dependence on a single partner into platformized, multi-model AI application revenue that serves global enterprise clients.