Amazon Web Services and OpenAI have expanded their partnership to make advanced AI models available within AWS’s existing cloud infrastructure. With this announcement, both parties are explicitly targeting large organizations that want to deploy AI without compromising on security, governance, and scalability.
Central to the partnership is the integration of OpenAI models into Amazon Bedrock, the AWS platform that allows companies to use various AI models through a single interface. This gives customers access to OpenAI’s latest models without leaving their existing AWS environment. The integration also means that organizations can continue to use their current security and compliance settings, including identity management, logging, and encryption.
Codex available via Bedrock
In addition to the models themselves, Codex will also be made available via Bedrock. This AI agent is designed for software development and can write, analyze, and improve code. Codex is already in widespread use and is intended to support development teams within AWS who want to automate and accelerate their workflows. Integration with tools such as Visual Studio Code makes it possible to use this functionality directly within existing development environments.
A third part of the announcement is the introduction of so-called Managed Agents within Bedrock, based on technology from OpenAI. These agents are designed to independently execute complex, multi-step processes, such as analyzing data or automating business processes. AWS handles a significant portion of the underlying infrastructure, including scalability, security, and monitoring. This should make it easier for organizations to actually deploy AI applications into production. According to AWS, these agents are specifically tailored to OpenAI’s reasoning models, with capabilities to control agent behavior and built-in security mechanisms.
According to Ben Kus, CTO of Box, companies are currently looking for ways to practically apply AI agents within their organizations. He states that the combination of OpenAI models and AWS infrastructure enables developers to build AI solutions that are not only scalable but also adapt to an organization’s specific context and meet governance and control requirements.
Microsoft Loses Exclusivity
The announcement follows shortly after a significant change in the relationship between OpenAI and Microsoft. Microsoft had long held exclusive rights to OpenAI technology within the cloud, but that exclusivity has been relinquished. This creates an opportunity for other cloud providers to offer OpenAI models. Amazon Web Services is responding immediately by accelerating the rollout of the integration into Bedrock.
CEO Andy Jassy responded shortly after the announcement of the revised agreement between OpenAI and Microsoft, stating that this development is significant, according to TechCrunch. The publication reports that this change enabled AWS to integrate OpenAI technology more broadly into its own services.
Increasing competition among AI partners
The collaboration between AWS and OpenAI is part of a broader strategic shift in the cloud and AI market. Whereas exclusive partnerships were previously the norm, companies are now explicitly seeking diversification. In addition to AWS, OpenAI also collaborates with Oracle, while Microsoft, in turn, is focusing on alternatives such as Anthropic and its Claude models, writes TechCrunch. This creates a more competitive playing field in which major technology companies encounter each other as both partners and competitors.
The scale of the collaboration underscores its strategic importance. The deal between Amazon and OpenAI is worth approximately $50 billion, indicating that both parties are committed to a long-term and capital-intensive partnership. For AWS, this strengthens its position as a platform for enterprise AI, while OpenAI expands its reach beyond Microsoft’s ecosystem.
The new features are currently available in a limited preview. It has not yet been announced when they will become more widely available. However, both parties indicate that this is a first step toward further deepening the collaboration, with the aim of bringing future AI developments to the AWS platform more quickly.