
The rift between Microsoft and OpenAI continues to grow and the changes they announced today are perhaps the biggest yet in the partnership.
“Since 2019, Microsoft and OpenAI have shared a vision to advance artificial intelligence responsibly and make its benefits broadly accessible,” the joint announcement explains. “What began as an investment in a research organization has grown into one of the most successful partnerships in our industry. As we enter the next phase of this partnership, we’ve signed a new definitive agreement that builds on our foundation, strengthens our partnership, and sets the stage for long-term success for both organizations.”
Here’s what’s changing.
Microsoft supports the OpenAI reorganization. With the trial between Elon Musk and OpenAI set to begin today, Microsoft announced its support of the OpenAI reorganization that will see the company change to a public benefit corporation (PBC). Microsoft will own about 27 percent of this company at a value today of $135 billion. Previously, Microsoft owned about 32.5 percent of OpenAI.
Frontier models. OpenAI will continue in its role as Microsoft’s frontier model partner, and Microsoft continues to have “exclusive IP [intellectual property] rights and Azure API exclusivity until [OpenAI achieves] Artificial General Intelligence (AGI).” AGI will now be declared by an independent expert panel.
Microsoft’s IP rights to OpenAI models and products. Microsoft previously had IP rights to OpenAI models and products through 2023; that has been extended to 2032 and now includes post-AGI models, which the two say will come with “appropriate safety guardrails.”
Microsoft’s IP rights to OpenAI research. Microsoft will continue to hold IP rights to OpenAI research—defined as “confidential methods used in the development of models and systems”—through 2030 or OpenAI achieving AGI, whichever comes first.
OpenAI consumer hardware. Microsoft no longer has IP rights to any consumer hardware that OpenAI releases in the future.
OpenAI joint development rights. OpenAI is now able to “jointly develop some products with third parties.” If these are API products, they will be exclusive to Azure. If they are non-API products, they can be served by any cloud provider.
Microsoft and AGI. Microsoft can now pursue AGI independently of OpenAI. It was sort-of doing this already, though it was describing the endgame as “superintelligence.” If Microsoft uses OpenAI’s IP to develop AGI, the models it makes will be “subject to compute thresholds; those thresholds are significantly larger than the size of systems used to train leading models today.”
Revenue share. The revenue share agreement between the two companies remains in place until either achieves AGI, but they’ve agreed to spread out those payments over an unspecified longer period of time.
OpenAI and Azure spending and exclusivity. OpenAI has agreed to spend an additional $250 billion on Azure services. In return “Microsoft will no longer have a right of first refusal to be OpenAI’s compute provider,” which is a big change. And OpenAI is now free to provide API access to US government national security customers, regardless of the cloud provider.
“As we step into this next chapter of our partnership, both companies are better positioned than ever to continue building great products that meet real-world needs, and create new opportunity for everyone and every business,” the announcement concludes.