Summary
Microsoft and OpenAI announced a renegotiated partnership on April 27, 2026 that ends Microsoft’s exclusive right to sell OpenAI’s models and stops Microsoft from paying revenue share on OpenAI products it resells through Azure. OpenAI will continue to pay revenue share to Microsoft through 2030, but that obligation is now subject to a total cap. OpenAI also gains the freedom to serve its products through any cloud provider, including AWS, Google, and Oracle. Microsoft retains a non-exclusive license to OpenAI IP through 2032 and remains a major shareholder, but the exclusivity that defined the partnership for nearly seven years is officially over.
Microsoft & OpenAI
Microsoft and OpenAI restructured their partnership again today, and this round is the clearest signal yet that the two companies are operating as competitors as much as collaborators.
The headline change: Microsoft will stop paying a revenue share to OpenAI on the AI products it resells through Azure. In exchange, OpenAI loses the exclusive cloud distribution arrangement that has tied it to Microsoft since 2019, and is now free to serve its full product lineup across AWS, Google Cloud, and Oracle.
It is the second major rewrite of the deal in seven months, and it lands less than a month after Microsoft released its own in-house AI models through Microsoft Foundry, an event we covered earlier this April.
What Actually Changed
The full set of changes was published in a joint statement and detailed in a Microsoft blog post.
Microsoft will no longer pay revenue share on OpenAI products sold through Azure. Until today, every paying Azure customer accessing OpenAI models triggered a payment from Microsoft to OpenAI. That flow has stopped.
OpenAI will continue to pay revenue share to Microsoft through 2030, at the same percentage as before, but the total amount is now capped. Microsoft does not have to wait for OpenAI to declare AGI for the obligation to end.
OpenAI’s exclusive cloud arrangement with Microsoft is over. OpenAI can now serve all of its products through any cloud provider. Microsoft remains the “primary cloud partner,” and OpenAI products will still ship first on Azure unless Microsoft decides otherwise, but exclusivity is gone.
Microsoft keeps a license to OpenAI’s IP and models through 2032, but that license is now non-exclusive. Microsoft remains a major shareholder, holding roughly 27% of OpenAI’s for-profit entity following its $135 billion valuation in October.
Why Now
The trigger was Amazon. In February, OpenAI signed a strategic partnership with AWS that included up to $50 billion in Amazon investment and an expanded $100 billion commitment to AWS infrastructure over the next eight years. AWS was named the exclusive third-party cloud distribution provider for OpenAI’s enterprise platform Frontier.
That deal created a problem under the old Microsoft contract, which gave Microsoft exclusive rights to OpenAI’s products until OpenAI internally declared AGI. The Amazon arrangement essentially needed today’s amendment to be legal.
OpenAI’s revenue chief Denise Dresser said in an internal memo earlier this month that the existing Microsoft partnership had “limited our ability to meet enterprises where they are.” Today’s deal removes that limitation.
Amazon CEO Andy Jassy confirmed on X that AWS will start serving OpenAI models through its Bedrock service in the coming weeks, with more details expected at an event in San Francisco tomorrow.
Why Microsoft Said Yes
Microsoft did not give up exclusivity for nothing. Three things sit on the other side of the trade.
First, Microsoft stops writing checks to OpenAI on every Azure resale. The exact size of those payments has not been disclosed, but Microsoft reported $7.5 billion in a single quarter from its OpenAI investment last quarter. Eliminating the outflow side of that flow is meaningful.
Second, Microsoft no longer has to model what happens when OpenAI declares AGI. The original agreement effectively gave OpenAI an off-ramp from its obligations the moment it claimed AGI, which created legal and strategic uncertainty Microsoft has now removed.
Third, Microsoft keeps the equity. Owning roughly 27% of OpenAI’s for-profit entity means Microsoft still benefits if OpenAI continues to grow, regardless of which cloud OpenAI sits on.
The trade is consistent with Microsoft’s broader posture over the last six months: build internal models, diversify partnerships including a recently expanded relationship with Anthropic, and treat OpenAI as one supplier in a portfolio rather than the supplier. The dynamic has been visible since GPT-5.4 launched in March, when OpenAI started accelerating product releases against an increasingly crowded field.
What This Means for the AI Market
The biggest immediate winner is AWS. Amazon now has a clear path to selling OpenAI models alongside Anthropic’s Claude through Bedrock, which gives it the most complete frontier model lineup of any major cloud. Google and Oracle are next in line.
OpenAI gets distribution flexibility at exactly the moment its enterprise platform is launching. Frontier was unveiled earlier this month, and the new deal lets OpenAI sell it across every major cloud without negotiating exceptions.
Microsoft gets simplification. The partnership had been sliding into open conflict, with both companies releasing competing models and going after the same enterprise accounts. Today’s deal acknowledges that reality and stops the partnership from being a fiction that constrains both sides.
The AI cloud market is now structurally different. Customers that previously had to choose between Azure for OpenAI access and AWS for Anthropic access can now run both from either cloud. Pricing pressure on cloud providers will increase, and the competitive moat for hosting frontier models has effectively been eliminated.
What It Means for Marketers
The model layer is fragmenting fast. ChatGPT, Claude, Gemini, and Copilot are no longer locked to specific clouds, distribution channels, or strategic partners. That has real consequences for AEO and AI search strategy.
Citation patterns will shift as OpenAI products show up on more clouds and integrate with more enterprise tools. Brands optimizing for ChatGPT visibility need to expect that the surface area where OpenAI models cite content is going to expand quickly into AWS, Google Cloud, and Oracle environments.
Cross-platform AEO becomes more important, not less. The brands that built strategies assuming a tight Microsoft-OpenAI corridor are now optimizing for a corridor that no longer exists. The brands that built for citation across multiple AI platforms are positioned to benefit. If you have not run a content gap analysis across the major AI engines yet, this is the moment.
For a deeper look at where AI search is heading and what that means for marketing strategy, our coverage of the future of search walks through the bigger picture.
The Bigger Shift
The Microsoft-OpenAI partnership defined the first phase of the AI boom. Microsoft’s $13 billion in investment, the exclusive cloud arrangement, and the integration of GPT into Copilot, Word, and Excel were the entire commercial story of generative AI for nearly three years.
That story is over. Today’s deal does not break the partnership, it formalizes the fact that the two companies are now running parallel, overlapping, sometimes competing AI businesses. Microsoft is building its own frontier models. OpenAI is selling its products through Microsoft’s biggest cloud rivals. The contract has caught up with what was already happening.
What comes next is a more open, more competitive, more fragmented model market. The brands and marketers paying attention now have a window to figure out what citation, visibility, and authority look like in a world where no single AI platform owns the answer layer.
Prompt Insider covers AEO, AI search, and the strategic shifts reshaping how brands earn visibility across ChatGPT, Claude, Gemini, and Perplexity. For a tactical breakdown on one of the highest-leverage AEO assets right now, see our latest on writing press releases that get cited by AI.