OpenAI and Amazon have just announced the official integration of OpenAI’s AI into AWS cloud infrastructure through the Amazon Bedrock service, further strengthening the strategic alliance the two companies announced in late February.
The launch includes three new limited preview activations in Amazon Bedrock: availability of OpenAI models, including GPT-5.5, in Bedrock; availability of Codex in Bedrock; and a new tool that makes it possible to create and manage agents in Amazon Bedrock powered by OpenAI technology.
This news comes just days after we learned that Microsoft and OpenAI have “opened up” their relationship, reshaping their agreement and leaving Sam Altman’s company free to offer its infrastructure and models through other companies such as Amazon.
“Microsoft remains OpenAI’s primary cloud partner, and OpenAI products will launch first on Azure, unless Microsoft is unable or chooses not to offer the required functionality. Now, OpenAI can offer all of its products to customers of any cloud provider,” OpenAI explained in its statement.
The expansion of the agreement between these two tech giants will allow AWS customers to use OpenAI models to deploy them in their processes and move forward with their development projects. They will even be able to orchestrate them alongside models from other providers such as Anthropic, Meta, Mistral, and Amazon itself, while maintaining the security controls, identity systems, and procurement processes they already trust.
As OpenAI states: “For developers, this means greater flexibility in how they build with OpenAI, from new AI applications to intelligence embedded in existing products, as well as agentic workflows that can reason, act, and support more complex business processes.
For businesses, this means a single, clear path from experimentation to production, with OpenAI capabilities available in the AWS environments where their most important workloads already run.”
Another major development in the agreement is the addition of Codex, OpenAI’s software development-focused model, to the Amazon Bedrock platform in limited preview. This integration allows companies to run AI-powered engineering processes within the infrastructure where their services already operate.
At present, more than 4 million professionals use Codex every week for task automation, code refactoring, and test generation, speeding up enterprise software delivery times. Its deployment on AWS through Bedrock will therefore provide a fresh boost for this already successful product, while also strengthening the cloud platform.
Availability in Bedrock ensures that technical teams use their AWS credentials for authentication and process inference through Amazon’s scalable infrastructure. In addition, the cost of the service is included within contracted AWS spending commitments.
“Codex in Bedrock is available through the Bedrock API, starting with the Codex CLI, the Codex desktop application, and the extension for Visual Studio Code,” Amazon explains.
The third strategic development in the OpenAI-AWS agreement is Amazon Bedrock Managed Agents, a tool that allows organizations to develop and deploy advanced AI agents, ready for production environments, using Amazon’s cloud infrastructure. Like the other two developments, it is available in limited preview.
By integrating OpenAI’s reasoning technology with AWS’s security and global scale, companies can automate complex tasks without having to manually manage the underlying infrastructure. The service is built on OpenAI’s agents framework to deliver optimized technical execution, precise reasoning, and reliable management of long-running processes.
From a security and governance standpoint, each agent operates with a unique identity and generates a detailed log of all its actions to support auditing processes. All model inference takes place entirely within Amazon Bedrock, ensuring privacy and proximity to corporate data and applications.
Amazon Bedrock Managed Agents simplifies the orchestration and use of AI tools, allowing technical teams to focus exclusively on the agent’s utility and functions. This integration ensures that scaling these systems across the enterprise complies with the operational and regulatory compliance standards that organizations already apply in their AWS environments.
Photo: generated with Nano Banana 2 (Gemini)
Your email address will not be published. Required fields are marked *
Δ