Headline: OpenAI lifts Azure-only tie with Microsoft — its models are headed to AWS and potentially Google Cloud
OpenAI and Microsoft announced on April 27 a major overhaul of their landmark 2019 partnership: Microsoft’s cloud license for OpenAI’s models is no longer exclusive. The move clears the way for OpenAI to sell its full model suite on Amazon Web Services and, potentially, Google Cloud — with Amazon CEO Andy Jassy confirming OpenAI models will arrive on AWS Bedrock within weeks.
What changed
- The April 27 joint statement converts Microsoft’s license from exclusive to non‑exclusive through 2032, effectively ending seven years of Azure-only distribution for OpenAI’s technology.
- OpenAI’s models (including the Codex agent) are already available to AWS customers through Amazon Bedrock; Amazon has introduced Bedrock Managed Agents powered by OpenAI so enterprises can build autonomous agents on AWS infrastructure.
- Amazon CEO Andy Jassy and AWS CEO Matt Garman have signaled strong customer demand; Garman said at an April 28 launch event that “this is what our customers have been asking us for for a really long time,” and OpenAI’s chief revenue officer Denise Dresser told staff the Microsoft arrangement had “limited our ability to meet enterprises where they are,” calling inbound demand for AWS “frankly staggering.”
Financial and legal details
- The restructuring ends the legal uncertainty that followed OpenAI’s February deal with Amazon, which had granted AWS exclusive third‑party cloud distribution for Frontier, OpenAI’s enterprise agent platform. Microsoft had been considering legal action; the April 27 change removes that overhang by explicitly making the license non‑exclusive.
- Microsoft still holds roughly 27% of OpenAI’s for‑profit arm and reported about $7.5 billion in revenue tied to that stake last quarter.
- Under the revised terms, Microsoft “no longer shares revenue with OpenAI,” while OpenAI will continue payments to Microsoft through 2030.
Practical impact for enterprises and cloud competition
- The most immediate effect: enterprises that previously needed Azure to run OpenAI models can now deploy those workloads on AWS and eventually Google Cloud, expanding choice and potentially lowering friction for multi‑cloud and cloud‑native projects.
- Google Cloud is reviewing the new deal terms to determine what kind of partnership might be possible under the non‑exclusive structure, Reuters reports.
- The two companies had been moving toward this new balance since June 2025, as competing product lines — from GitHub Copilot vs. OpenAI’s Windsurf to Microsoft’s own LLM efforts — made exclusivity increasingly untenable.
Why it matters for the broader tech and crypto ecosystems
- More cloud options for OpenAI models could accelerate adoption among enterprises, independent developers and blockchain/web3 teams that rely on large‑scale cloud infrastructure for model hosting, agent orchestration and integrated services.
- Greater distribution and multi‑cloud availability may spur new integrations, lower vendor lock‑in, and intensify competition among Azure, AWS and Google Cloud for AI workloads — outcomes with downstream effects for startups and crypto projects that build on AI-powered tooling.
What to watch next
- AWS Bedrock rollouts in the coming weeks and any formal Google Cloud partnership announcement.
- How pricing, revenue sharing and technical integrations (APIs, managed agents, compliance tooling) shape enterprise adoption.
- Any further legal or governance clarifications tied to Microsoft’s remaining stake and the payment terms through 2030.
Bottom line: The move away from an Azure‑only model opens OpenAI’s stack to a broader cloud market, reshuffling competitive dynamics between hyperscalers and giving enterprises and developers more choice in where they build with OpenAI technology.
Read more AI-generated news on: undefined/news
