OpenAI ends its exclusive partnership with Microsoft
OpenAI breaks exclusivity with Microsoft, bringing models to AWS and others.
In a significant shift for the AI industry, OpenAI and Microsoft have announced an amended agreement that ends the exclusive cloud partnership established in 2019. Under the new terms, OpenAI can now offer its models through any cloud provider, including Amazon Web Services (AWS) and Google Cloud, breaking Microsoft's previous monopoly. The deal clarifies that Microsoft retains a non-exclusive license for OpenAI's intellectual property and models through 2032, with Azure remaining the primary cloud partner as long as Microsoft can honor that commitment.
This amendment comes two months after a $50 billion deal between OpenAI and Amazon, which included plans for OpenAI models to run on AWS's Bedrock service. The Financial Times reported that Microsoft had threatened legal action over that deal, but today's announcement moots those concerns. OpenAI Chief Revenue Officer Denise Dresser noted in a memo that the Microsoft partnership had limited the company's ability to meet enterprise demand, particularly on Amazon Bedrock. Amazon CEO Andy Jassy expressed excitement about making OpenAI models available on Bedrock in the coming weeks, alongside a new Stateful Runtime Environment. The revenue share from OpenAI to Microsoft, previously 20%, is now capped and runs through 2030, independent of any AGI milestones.
- OpenAI can now offer models on any cloud provider, ending Microsoft's exclusive Azure deal.
- Microsoft retains a non-exclusive license for OpenAI's IP through 2032, with Azure as primary partner.
- Revenue share is capped and runs through 2030, removing the AGI clause that previously tied exclusivity to AGI achievement.
Why It Matters
This opens the AI market to multi-cloud competition, giving enterprises more flexibility and choice.