OpenAI deal capped at $38B

OpenAI’s Microsoft deal just got a lot less open-ended. The new wrinkle is a number — $38 billion — which is now the ceiling on how much OpenAI will pay Microsoft in revenue share through 2030. That lands a couple of weeks after the bigger reset on April 27, when the two companies loosened one of the most important partnerships in tech and ended Microsoft’s exclusive hold on OpenAI distribution. ### What actually changed? The old setup tied OpenAI tightly to Microsoft. Microsoft had exclusive rights around OpenAI’s IP and commercial distribution in ways that made Azure the default home for OpenAI products. The amended deal keeps Microsoft as OpenAI’s primary cloud partner, but not the only one, and Microsoft’s license to OpenAI model IP now runs through 2032 on a non-exclusive basis. (money.usnews.com) ### Why is the $38 billion cap a big deal? Because “subject to a cap” and “the cap is $38 billion” are very different things. In April, OpenAI and Microsoft said the revenue share would continue through 2030 at the same percentage as before, but they did not publish the ceiling. The newer reporting fills in the missing number — $38 billion — and says the rewrite could save OpenAI about $97 billion through 2030 versus the older path. That turns a vague future obligation into something investors can model. (blogs.microsoft.com) ### Did Microsoft give up a lot? Yes, but not everything. Microsoft no longer pays revenue share to OpenAI, and it no longer holds an exclusive license to OpenAI technology. But Microsoft still keeps a major equity stake, still gets revenue-share payments from OpenAI through 2030, and still remains the primary cloud partner. Basically, Microsoft gave up control points that boxed OpenAI in, while keeping the economics and strategic access that still matter. (blogs.microsoft.com) ### Why does cloud exclusivity matter so much? Because enterprise customers do not all live on Azure. OpenAI had been constrained by a setup that pushed its products through Microsoft’s stack first. The new terms let OpenAI serve products across other providers, including AWS and Google, which is a much more normal way to sell infrastructure-heavy software. And the change was not theoretical — OpenAI models showed up on AWS Bedrock almost immediately after exclusivity ended. (blogs.microsoft.com) ### What happened to the AGI clause? It appears to be gone from the revised arrangement. That clause had long been one of the strangest parts of the Microsoft-OpenAI relationship because it created a moving boundary around what Microsoft could access if OpenAI declared it had reached AGI. Removing it makes the contract simpler and more legible. The catch is that neither company’s public post made that point the centerpiece — it emerged mainly through follow-on reporting. (cnbc.com) ### Is this really about an IPO? Not officially, but that is the obvious inference. A company heading toward public markets wants fewer weird related-party obligations, cleaner revenue rights, and less dependence on one distributor. Capping the Microsoft payout, ending exclusivity, and simplifying licensing all move in that direction. Several follow-on analyses tied the rewrite directly to making OpenAI easier to value as a future public company. (winbuzzer.com) ### So what’s the bottom line? This is not a breakup. It is a rewiring. OpenAI gets more room to sell, more freedom to choose infrastructure, and a hard stop on a giant payment stream. Microsoft loses exclusivity, but keeps deep commercial ties and upside. In plain English — the partnership just shifted from “near-merger with escape hatches” to “big investor and preferred partner, but not the landlord.” (blogs.microsoft.com) (augment.market)