OpenAI and Microsoft have agreed to cap revenue-sharing between them at $38 billion, a move that signals the AI company’s growing push for independence and could upend the cloud computing landscape. The deal, which sets a limit on how much money Microsoft can take from OpenAI’s revenue, comes as the startup looks to loosen its reliance on its biggest backer. The cap effectively puts a ceiling on the financial relationship that has defined the generative AI boom, and it’s already prompting questions about who will control the next wave of artificial intelligence.
Why the $38 billion cap matters
The revenue-sharing cap is not a one-time payment or an investment round — it’s a structural change to the partnership. Under the previous arrangement, Microsoft received a cut of OpenAI’s revenue, which grew rapidly as ChatGPT and other products took off. Now that cut has a hard limit. For Microsoft, the cap means it will no longer see unlimited upside from OpenAI’s success. For OpenAI, the cap is a step toward financial independence: once the $38 billion threshold is reached, the company keeps every additional dollar it earns. That changes the incentive structure for both sides. OpenAI has been on a fundraising spree, and this cap makes it clearer to investors that the company won’t be permanently tied to Microsoft’s income stream.
OpenAI’s independence drive
OpenAI has been quietly moving away from Microsoft for months. The company has built its own cloud infrastructure, hired top talent from rival firms, and explored partnerships with other cloud providers. The revenue-sharing cap formalizes that ambition. OpenAI wants to control its own destiny — its own data, its own compute, its own revenue. The $38 billion figure wasn't pulled from thin air; it reflects the existing revenue streams from Microsoft-OpenAI joint products and serves as a bridge to full independence. While Microsoft remains OpenAI’s preferred cloud partner, the cap reduces the risk of Microsoft one day owning the AI company’s economics entirely. It’s a delicate balance: OpenAI still needs Microsoft’s Azure capacity to train and run its models, but the cap means it won’t be forever splitting the profits.
The cap could reshape cloud dynamics. Microsoft has leveraged its OpenAI partnership to pull ahead in the cloud AI race against AWS and Google Cloud. If OpenAI is no longer a permanent profit-sharing partner, Microsoft may have to compete harder for OpenAI’s business — or for other AI startups’ business. Amazon and Google have already poured billions into their own AI models, and the cap gives them an opening. They can now argue that OpenAI isn’t locked into Microsoft forever. For the broader AI market, the cap signals that even the most successful AI startup doesn’t want to be owned by a cloud giant. That could encourage other AI companies to negotiate similar caps, potentially fragmenting the market. The power structure of AI may shift as a result: fewer companies will hold all the cards, and more startups may feel emboldened to demand independence from their cloud providers.
What remains unanswered is how quickly OpenAI can hit that $38 billion ceiling — and what happens to the partnership once it does. Microsoft has said little publicly about the change, and OpenAI hasn’t detailed its timeline. The next earnings calls from both companies will likely offer the first real clues.
