OpenAI & Microsoft Partnership Is On Shaky Ground as Altman Admits ‘Points of Tension’
In a New York Times podcast released late Tuesday, OpenAI CEO Sam Altman revealed he had a private call with Microsoft chief Satya Nadella “on Monday” to hash out the future of their alliance, conceding there are “points of tension.”
Microsoft is OpenAI’s biggest investor. Back in 2019, Microsoft wrote a $1 billion check for OpenAI, and in return, gained preferred access to OpenAI’s models. That stake has since ballooned to an estimated $13 billion.
Parties disagree on Microsoft’s future stake in OpenAI
“The points of tension” Altman refers to are concerning both parties’ disagreements on Microsoft’s shares and earned stance in OpenAI’s future. The contract Microsoft and OpenAI made in 2019 will legally end in 2030.
Microsoft is pushing for a larger equity cut than the 33% OpenAI has proposed for the renewal, along with exclusive rights to resell OpenAI APIs on Azure and influence over OpenAI’s planned $3 billion acquisition of Windsurf, a direct competitor to GitHub Copilot; demands that OpenAI is rejecting.
On the other side, OpenAI is pushing to reduce the 20% revenue share it currently pays Microsoft, maintain control over Windsurf and its own intellectual property, and preserve the freedom to run future models on non-Azure cloud platforms like Oracle, Google, or possibly even a Stargate-backed data center.
OpenAI also wants to turn itself into an ordinary profit-making company and raise $40 billion from investors led by SoftBank. Due to the nature of their partnership, OpenAI needs Microsoft’s approval to be able to do this.
How tense are things?
During the New York Times podcast last Tuesday, Altman stated that even though there are points of tension, the collaboration has been “wonderfully good” overall.
However, a few weeks ago, the ChatGPT creator was evaluating whether they should seek a federal regulatory review of the terms of its contract with Microsoft for potential violations of antitrust law, as well as a public campaign, because Microsoft is acting anticompetitively in their partnership.
What does it mean for the end user?
While the debates are heated, it is highly unlikely for the end user to see the impact overnight. The parties still have time to come to some sort of agreement. If they succeed, nothing will change for the end user.
If they fail, many things can happen.
Reuters reported that Microsoft plans to exploit its existing contract with OpenAI to access its technology until 2030, without a hint of trying to agree with the ChatGPT creator to extend the agreement. If Microsoft continues with its plans, OpenAI could lose billions in new funding, and Microsoft might freeze extra cash, which could reflect on both companies’ shares and value.
On the other hand, OpenAI has been in talks with SoftBank and Oracle about a $500 billion data-center project called “Stargate.” If Microsoft decides to walk away, OpenAI could run future AI models on Stargate instead of on Microsoft’s Azure cloud.
In the meantime, the US FTC and EU regulators may take action, depending on the outcome of the discussions. The regulators have already been closely monitoring the OpenAI and Microsoft partnership.
While the FTC describes this union as “a cloud-AI partnership that could harm competition”, the EU has been examining whether Microsoft’s stake in OpenAI should be treated like an acquisition under EU merger rules. In the event of a failure to agree, regulators may take action against both companies.