Executive Summary:
- Microsoft disclosures and Bloomberg analyses show that OpenAI's compute spend and revenue share accounted for 70% or more of Microsoft's FY26 AI revenues, and more than 7% of Microsoft's overall FY2026 revenues.
- Microsoft has spent $261.3 billion in capital expenditures since the beginning of 2022.
- OpenAI accounted for $24.1 billion of Microsoft's FY2026 revenues.
As I discussed in yesterday's free newsletter, analyst estimates have OpenAI and Anthropic making up over 70% of all AI revenues across Microsoft, Google and Amazon.
While some might have disagreed, Bloomberg is now reporting that OpenAI "accounted for more than half, and likely about 70%, of Microsoft's actual AI sales during its most recent fiscal year."
Bloomberg's maths is explained as such:
"Bloomberg’s analysis assumes that Microsoft’s annual AI run rate continued growing at the rapid 123% the company reported for March. At that growth rate, the company’s AI business would have tallied roughly $34 billion in the fiscal year ending in June, which can be directly compared with the new disclosure that OpenAI provided $24.1 billion of revenue in that year."
The actual disclosure from Microsoft comes from its most-recent earnings:
For fiscal year 2026, we recorded revenue from commercial arrangements with OpenAI, inclusive of revenue sharing payments, of $24.1 billion, and accounts receivable from OpenAI as of June 30, 2026 was $6.0 billion.
To be clear, "run rate" means a non-specific month multiplied by 12, which means that it's very possible that Microsoft actually made far less than $37 billion, but based on that maths, OpenAI would make up roughly 64.8% - that being said, I think 70% or more is a perfectly-reasonable estimate, if not far, far more.
The other incredible fact from these disclosures is that OpenAI's spend and revenue share accounted for 7% of Microsoft's $331.8 billion in FY26 revenue - or around 7.26% to be specific.
At this point, it's impossible to argue that Microsoft has spent $270 billion in capital expenditures to prop up a single client, and that its overall AI plays have failed to create any significant revenue growth or opportunities.
We are now four years into the AI bubble, and Microsoft has little to show for it other than one very large and very unsustainable company that requires near-infinite resources to keep paying its cloud compute bills.
If you liked this piece, you should subscribe to my premium newsletter. It’s $70 a year, or $7 a month, and in return you get a weekly newsletter that’s usually anywhere from 5,000 to 18,000 words, including vast, detailed analyses of NVIDIA, Anthropic and OpenAI’s finances, and the AI bubble writ large. My Hater's Guides To the SaaSpocalypse, Private Credit and Private Equity are essential to understanding our current financial system, and my guide to how OpenAI Kills Oracle pairs nicely with my Hater's Guide To Oracle, as well as the Hater’s Guide To Oracle (Part 2).
Subscribing to premium is both great value and makes it possible to write these large, deeply-researched free pieces every week. On Friday, I’ll publish the second installment of the Hater’s Guide to Nvidia — where I’ll take a look at how the AI bubble transformed the company from a pure hardware player to a purveyor of the financial dark arts.
If you want to get in touch — and especially if you have any juicy information about Anthropic, OpenAI, or any other companies in the AI bubble — hit me up on Signal at ezitron.76. I’m also on IB on The Terminal.
