What Microsoft hopes to get from its breakup with OpenAI

The once-tight bond between Microsoft and OpenAI has been fraying for well over a year — and it’s getting worse. The companies were at one point so close that OpenAI CEO Sam Altman called their relationship “the best bromance in tech.” 

But as in many romances, times change. Relationships sour. 

For at least six months, the two companies have been sniping at each other. At times, there’s been open warfare. Altman tried to convince more than 300 corporate executives around the world to switch their AI allegiance from Microsoft to OpenAI by telling them they’d be better off working directly with the people who built generative AI (genAI) technology rather than getting it secondhand via Copilot from Microsoft.

Microsoft CEO Satya Nadella dissed OpenAI, saying bluntly: “If OpenAI disappeared tomorrow…, we have all the IP rights and all the capability. We have the people, we have the compute, we have the data, we have everything. We are below them, above them, around them.”

Now, it’s time for the lawyering up. You can easily guess what the fight is over: What Microsoft will ultimately get for its $13 billion investment in OpenAI and what their relationship looks like in the long run.

The roots of the fight

The confusion about what Microsoft gets in return for its big OpenAI investment has its roots in the way the company was founded in 2015. It originally started as a nonprofit. Founders, including Altman, Elon Musk, and others, worried that if AI were allowed to be developed unchecked, it could become a threat to humanity. They funded the company out of their own pockets, and said the company’s purpose would be to ensure that AI would be “used in the way that is most likely to benefit humanity as a whole, unconstrained by a need to generate financial return.”

It’s easy to have high-minded goals like that when little or no money is at stake. But eventually reality intervened in the form of the hundreds of trillion dollars of potential AI profits. Altruism disappeared and the founders turned OpenAI into a “capped” for-profit company, with a nearly incomprehensible corporate structure. OpenAI opened itself up to investors. More confusing still, OpenAI was allowed to compete directly with its investors.

Along the way, Microsoft invested in the company and was granted the sole rights to use OpenAI technologies, which it used to build its Copilot AI tools. But it’s never been clear how Microsoft could cash in over the long run.  

Since then OpenAI has outlined a complicated plan for restructuring that would allow it to launch an IPO — and it looks as if that plan will eventually proceed. 

 That brings us to today.

The rubber meets the road

Before OpenAI can go public via an IPO, it needs to come to an agreement with Microsoft about what Microsoft’s investment will be worth. Both companies are very different today than when that investment was made. OpenAI is now a big and successful operation, valued at $300 billion, with arguably the most advanced AI capabilities in the world.

Microsoft has built itself into a $1 trillion+ company on the back of using OpenAI’s ChatGPT to underpin its Copilot AI tools, which are now embedded in every part of the company. In addition, Microsoft’s hosting of AI services appears to cement its lead as the world’s largest AI company. What they want from each other is now very different than when Microsoft made its original $13 billion investment.

These days, Microsoft is less interested in getting a big cash infusion from an IPO than in getting long-term access to OpenAI technologies that can help it maintain and expand its AI lead.

“Microsoft’s interests go beyond equity — it’s about return on investment, sustained AI infrastructure growth, and maintaining Copilot’s edge by anchoring it on OpenAI models,” said Neil Shah, vice president for research and partner at Counterpoint Research. “Microsoft would like to continue having a big slice of it, being one of the key initial supporters and investors.”

“If Microsoft trades equity for extended model access, it’s signaling a pragmatic pivot from ownership to operational leverage,” said Abhivyakti Sengar, practice director at Everest Group. “This would give Microsoft continued dominance across enterprise applications without the burden of influencing OpenAI’s increasingly complex governance structure.”

As to what OpenAI wants from any agreement, that’s a little less clear, in part because it claims to be pursuing two seemingly contradictory goals: Making sure AI is developed ethically, while making as much profit as possible. 

It’s also possible that its reorganization for an IPO might not pass muster with regulators. So, it’s in the company’s interest to come to a friendly deal with Microsoft, which would make it more likely regulators will approve the new structure.

This means that Microsoft is more likely than not to get what it wants from OpenAI — long-term access to the company’s most valuable technologies, which are worth far more to the company than cash. If that happens, it’s hard to imagine anyone toppling Microsoft as the world’s most valuable AI company.