Microsoft Corp. (MSFT) generates the vast majority of its rapidly growing artificial intelligence revenue from a single source: OpenAI, according to a regulatory filing that sheds new light on the deep financial entanglement between the software giant and the creator of ChatGPT.
In a disclosure made last week, Microsoft revealed that it recorded $24.1 billion in revenue from OpenAI during the fiscal year ended June 2026. The figure, which includes payments for cloud computing power, model development costs, and a share of OpenAI’s revenue, dwarfs the company’s previously stated AI sales benchmarks and underscores a dependency that has drawn increasing scrutiny from Wall Street.
Microsoft Chief Executive Officer Satya Nadella had previously stated that the company’s AI business was on an annualized revenue run rate exceeding $37 billion as of the March quarter. While Microsoft did not update that total figure alongside its fiscal fourth-quarter earnings report, a Bloomberg analysis suggests that OpenAI likely accounted for roughly 70% of Microsoft’s actual AI sales for the full fiscal year, assuming the triple-digit growth rate persisted.
“The disclosure suggests that OpenAI accounted for more than half, and likely about 70%, of Microsoft’s actual AI sales during its most recent fiscal year,” the analysis indicates, highlighting how dependent Microsoft remains on its close partner.
A Microsoft spokesperson confirmed that the total AI business figure includes all sales and revenue share from OpenAI. When measured against Microsoft’s overall revenue, however, OpenAI’s contribution represents less than 10% of the total, a much smaller slice of the company’s sprawling enterprise empire.
The revelation comes as Microsoft is simultaneously attempting to diversify its AI portfolio and manage internal costs. In a separate memo viewed by CNBC, Jay Parikh, executive vice president of Microsoft’s CoreAI engineering group, instructed developers to default to OpenAI’s flagship GPT-5.6 Sol model when working in the GitHub Copilot coding tool. The directive is part of an efficiency push to maximize the value of Microsoft’s massive token investments.
“Internally, shifting more workloads to OpenAI models helps us get greater value from our token investment,” Parikh wrote in the memo. He noted that while divisions are managing AI token budgets, individual teams have not yet been assigned specific spending limits, though he encouraged employees to discuss large-scale projects with their managers.
The internal guidance arrives nine months after OpenAI completed a corporate restructuring that extended intellectual property rights for Microsoft through 2032. Microsoft said in April it was ceasing revenue-sharing payments to OpenAI, a move that came as the company deepened its ties with rival AI lab Anthropic PBC, agreeing to invest up to $5 billion and releasing the Copilot Cowork product containing its models.
Anthropic committed to spend $30 billion on Microsoft Azure cloud services, but the companies have not made public assurances on intellectual property comparable to the OpenAI deal.
Olga Usvyatsky, an accounting researcher and founder of data-analytics firm Nonlinear Analytics, suggested the timing of the OpenAI revenue disclosure may be tied to the startup’s plans for an initial public offering. The filing marks the first time Microsoft has clearly released its full revenue from the AI firm.
For investors, the critical unanswered question revolves around the quality of that revenue stream. KeyBanc analyst Jackson Ader said it remains unclear how much of OpenAI’s contribution stems from the revenue-sharing agreement versus traditional cloud computing or other services.
“The more of that revenue comes from services to OpenAI rather than the benefits of investment, the more favorably I’m going to look at it,” Ader said.
The financial dependency narrative is unfolding against a backdrop of tightening capital discipline across the technology industry. Wall Street is demanding greater returns from the hyperscalers’ massive AI spending commitments, with capital expenditures from Microsoft, Amazon, Alphabet, and Meta expected to collectively top $700 billion this year. Free cash flow across the group dwindled in the latest quarter, though Microsoft’s 23% decline was relatively mild compared to its peers.
Microsoft’s stock rallied 22% last week on earnings, its strongest weekly performance since 1999, though it remains up only about 1% for the year, trailing most megacap rivals. The company added about $51 billion in commercial bookings in its most recent quarter, with OpenAI making up the majority of annual bookings growth.
The CoreAI group, which includes GitHub, Visual Studio, and Visual Studio Code, will continue to adjust default model settings as products evolve, according to Parikh’s memo. A Microsoft spokesperson said the company periodically updates “the default model settings in our internal tools to balance performance and efficient use of resources,” while continuing to offer a range of model options that engineers can select at any time.