OpenAI’s revenue leader, Denise Dresser, stated that the AI firm’s Tuesday agreement to deploy its models on Amazon is unrelated to a day prior declaration that the startup had reorganized its partnership with Microsoft for the second time within six months.

«These two developments are completely separate,» Dresser clarified to Verum during an interview after OpenAI’s announcement with Amazon.

However, market analysts remain skeptical.

Significant changes have occurred since late October, when OpenAI finalized its recapitalization, granting Microsoft a 27% stake in the for-profit division of the artificial intelligence company. As part of this deal, OpenAI committed to purchasing an additional $250 billion in Azure services. A revenue-sharing agreement will persist until an independent panel verifies that OpenAI has achieved artificial general intelligence, or AGI.

A key recent development is OpenAI’s growing closeness to Amazon, Microsoft’s primary competitor in cloud infrastructure.

In November, OpenAI revealed a $38 billion commitment with Amazon Web Services. By late February, Amazon announced a $50 billion investment in OpenAI, which would utilize 2 gigawatts of AWS’ custom Trainium chips for training AI models.

Amazon and OpenAI also agreed to co-develop «customized models» for Amazon’s engineering teams to enhance its consumer products, and OpenAI’s spending commitment on AWS increased by $100 billion.

«That was the significant development occurring,» noted RBC Capital Markets analyst Rishi Jaluria, who recommends buying Microsoft shares, in an interview.

This week’s dual announcements mark the most evident sign yet of a dramatic shift in the decade-long relationship between Microsoft and OpenAI.

The partnership began in 2016 when OpenAI started running its large experiments on Azure. Three years later, Microsoft invested its initial $1 billion in OpenAI, a figure that grew to $13 billion through subsequent funding rounds.

However, in 2024, Microsoft began labeling OpenAI as a competitor in its financial reports, and early last year, the software giant lost its status as OpenAI’s exclusive cloud provider. In an internal memo earlier this month, Dresser wrote that OpenAI’s partnership with Microsoft has been «foundational to our success,» but «has also limited our ability to meet enterprises where they are.»

Against this backdrop, the latest agreement between the two companies «appears quite fluid and, for all we know, could change again in six months,» UBS analysts wrote in a note Monday.

Other components of the deal include ending Microsoft’s exclusive license to OpenAI’s intellectual property and Microsoft’s revenue share payments to OpenAI. Microsoft will also no longer be the sole cloud provider for API products built with third parties.

«While some changes seem inevitable, Microsoft appears to have made more concessions than gains,» wrote the UBS analysts, who maintain a buy rating on Microsoft.

Amazon CEO Andy Jassy called Monday’s announcement «very interesting» in a post on X, adding that more details would be shared Tuesday.

Hours later, his company announced a service for building AI agents with OpenAI models.

‘Original partner’

For years, developers interested in those models needed to go through Microsoft’s Azure cloud or work with OpenAI directly. Now, companies with large AWS investments will be able to more easily adopt the models, while taking advantage of volume spending plans.

Dresser, speaking from an Amazon event, said the reworking of OpenAI’s arrangement with Microsoft was not inspired by the growing collaboration with Amazon.

«Microsoft is our original partner,» she said. «They’re an incredible partner to us. They will be a premier partner as we move forward. What we are focused on is making sure, as we meet our customers where they are, that they have access to environments that they’re working in. And we want to make sure that we deliver the best models in the best environments for customers to be successful.»

The Financial Times reported that Microsoft considered legal action regarding OpenAI’s plans with Amazon, and Microsoft told the newspaper that it was «confident that OpenAI understands and respects the importance of living up to [its] legal obligation.» Microsoft didn’t provide a comment beyond Monday’s announcement.

Microsoft is similarly making moves to diversify away from OpenAI.

In September, Microsoft said it was starting to draw on an AI model from Anthropic to answer some queries in the 365 Copilot assistant for commercial clients. Two months later, Microsoft agreed to invest up to $5 billion into Anthropic, which committed to purchasing $30 billion of Azure compute capacity.

Taking advantage of the surging popularity of Anthropic’s Claude Code, Microsoft announced in March an offering called Copilot Cowork in cooperation with Anthropic.

One downside of soaring demand for Claude is that reliability has suffered. The company reported partial or major outages during 37 of the past 90 days. Amazon, an early Anthropic partner and investor, has taken notice.

Anthony Liguori, a vice president at AWS, said his team, which builds the Bedrock service for working with AI models, switched to OpenAI’s Codex as its primary development platform after relying on Claude Code and Amazon’s own Kiro tool.

The reality for all the major parties involved is that they need each other.

Capacity is so constrained that OpenAI and Anthropic need to work with all of the major cloud vendors to secure as much compute as possible. And Microsoft and Amazon need simple access to all the major models to serve their massive customer bases.

So while Microsoft and OpenAI may be drifting apart, Jaluria was quick to note, «Microsoft still needs OpenAI, and OpenAI still needs Microsoft.»

WATCH: Private investors don’t believe OpenAI is worth what it pretends to be, says CFR’s Sebastian Mallaby