OpenAI's models, Codex (its code-generation tool), and Managed Agents are now available directly on AWS. The announcement means businesses already running on Amazon's cloud can build AI-powered products without routing data through OpenAI's own infrastructure.
The key word here is "secure." Many enterprise customers have been hesitant to use OpenAI's APIs because it means sending sensitive data to a third-party endpoint outside their existing cloud agreements, compliance frameworks, and network boundaries. Running through AWS means that data can stay within an organization's existing AWS environment, subject to the same governance controls they already have in place.
Managed Agents - OpenAI's framework for building AI systems that can take sequences of actions, call tools, and complete multi-step tasks without constant human input - are the more interesting inclusion here. Hosting them on AWS gives developers a familiar deployment target and keeps the orchestration layer inside existing infrastructure. For teams already using AWS Lambda, ECS, or Bedrock, this removes a significant integration hurdle.
Codex, OpenAI's model built specifically for writing and editing code, being on AWS is a natural fit. Development teams often have strict policies about where source code can go. An AWS-native deployment path makes Codex more viable in regulated industries like finance and healthcare, where sending proprietary code to an external API would require extensive legal review.
This follows a pattern across the industry: Microsoft has Azure OpenAI Service, Google has Vertex AI with Gemini, and now OpenAI is making a direct push to be available in the one major cloud it wasn't deeply embedded in. For enterprises that standardized on AWS and were previously weighing Amazon Q Developer against building wrappers around OpenAI's APIs, this simplifies the decision considerably. OpenAI's full announcement is on their blog.