For years, Microsoft and OpenAI operated as something close to a single entity in the AI race. Microsoft poured roughly $13 billion into the startup. OpenAI ran almost exclusively on Microsoft’s Azure cloud. And Microsoft collected a revenue share on OpenAI’s commercial products, a financial umbilical cord that made every ChatGPT subscription partly a Microsoft payday.
That cord has been cut. Under an amended partnership agreement finalized in early 2025, Microsoft no longer receives a revenue share from OpenAI, according to reporting by the Associated Press. OpenAI, meanwhile, remains contractually committed to paying Microsoft for Azure cloud services through 2030. The restructuring handed OpenAI full control of its own product revenue for the first time since the partnership began, while Microsoft locked in years of guaranteed cloud fees.
Days later, Amazon made its move. The company announced a major expansion of its relationship with OpenAI, a deal reportedly valued at up to $50 billion in combined investment and cloud commitments. The timing was not subtle. Amazon had already committed roughly $8 billion to Anthropic, OpenAI’s chief rival, but the chance to court the most recognized name in generative AI proved too significant to pass up.
What the restructured deal actually changes
The financial mechanics matter here. Under the original arrangement, Microsoft earned money from OpenAI in two ways: cloud fees for hosting OpenAI’s models on Azure, and a percentage of OpenAI’s commercial revenue. Eliminating the revenue share removes the second stream but preserves the first, which is arguably more valuable as OpenAI’s computing needs continue to balloon.
For OpenAI, the trade is straightforward. The company keeps more of its own money at a moment when it is burning through cash to train increasingly expensive models. OpenAI’s annualized revenue reportedly surpassed $5 billion in late 2024, and the company has been pursuing a conversion from its unusual capped-profit structure to a full for-profit corporation. Shedding the revenue-share obligation strengthens OpenAI’s balance sheet ahead of that transition and makes the company more attractive to new investors and partners.
For Microsoft, the calculation is more layered. Giving up a direct cut of OpenAI’s revenue sounds like a concession, but Microsoft still embeds OpenAI’s models across its product line, from Copilot in Office to Azure AI services sold to enterprise customers. Those integrations generate revenue Microsoft keeps entirely. And the guaranteed cloud payments through 2030 provide a predictable income floor regardless of how OpenAI’s own business performs.
Amazon’s $50 billion opening
Amazon’s expanded partnership with OpenAI represents a significant escalation in the cloud wars over AI. The company had already established itself as Anthropic’s primary backer, but moving toward OpenAI signals a willingness to support multiple leading AI developers simultaneously.
The full terms of the Amazon-OpenAI expansion have not been publicly detailed. What is known is that Amazon moved with striking speed once the Microsoft-OpenAI restructuring became public, suggesting months of quiet groundwork. Whether the arrangement involves direct equity investment, cloud hosting agreements on Amazon Web Services, product integrations, or some combination remains to be confirmed.
For Anthropic, the development raises uncomfortable questions. Amazon’s billions helped Anthropic compete with OpenAI on model development and enterprise sales. If Amazon now divides its attention and resources between two rival AI labs, Anthropic could find itself sharing a benefactor rather than enjoying one exclusively.
Regulators are already watching
None of this is happening in a vacuum. The U.S. Federal Trade Commission launched a formal inquiry in January 2024 into generative AI investments and partnerships, demanding detailed information about agreements, strategic rationale, and governance structures from Microsoft, OpenAI, Amazon, Anthropic, and Google’s parent company Alphabet.
The inquiry has not yet produced public findings or enforcement actions as of June 2026. But its existence shaped the environment in which the Microsoft-OpenAI restructuring took place. Both companies had reason to loosen visible financial entanglements that regulators might characterize as anticompetitive. A partnership where one company both invests billions and collects a revenue share from another’s products is exactly the kind of arrangement the FTC flagged for scrutiny.
Whether the commission ultimately concludes these deals pose competitive harm is an open question. The FTC could issue new rules restricting preferential treatment between cloud providers and AI labs, negotiate consent orders with specific companies, or simply publish a report and move on. Until that direction becomes clear, every major AI partnership carries a layer of regulatory risk that did not exist two years ago.
What this signals for the AI industry
The era of exclusive, tightly bound alliances between cloud giants and AI startups appears to be ending. OpenAI’s willingness to restructure its most important corporate relationship, combined with Amazon’s eagerness to step in, points toward a more fluid landscape where AI developers play cloud providers against each other for better terms.
That shift has practical consequences for companies building products on top of AI models. If OpenAI’s technology becomes available across multiple cloud platforms rather than primarily through Azure, enterprise customers gain leverage on pricing and deployment flexibility. The same logic applies if Anthropic’s models eventually run on infrastructure beyond AWS.
Microsoft is not walking away from AI. The company still holds a significant stake in OpenAI, still integrates OpenAI’s models into its flagship products, and still collects cloud fees that will run for years. But the relationship has moved from something resembling a merger of interests to something closer to a commercial arrangement between two companies with increasingly independent ambitions.
The altered deal stands as a visible marker of that transition. It preserves a lucrative cloud pipeline for Microsoft, grants OpenAI direct control over its own revenue, and tells every competitor in the industry that the once-closed circle around the leading AI lab is now open for business. How Amazon, Anthropic, Google, and regulators respond will determine whether this moment is remembered as a tactical adjustment or the beginning of a broader realignment in how AI power is financed and distributed.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
