OpenAI Models and Codex Launch on Amazon Bedrock

OpenAI's frontier models and its Codex software engineering agent are now generally available on AWS, opening a direct path for millions of Amazon customers to build with OpenAI inside the platform they already run their businesses on. According to OpenAI's announcement, the rollout is designed to help enterprises move AI into production through the security, compliance, and governance workflows their teams already trust. For many organizations, that is the part that has held adoption back.

Both offerings run through Amazon Bedrock. OpenAI models on Amazon Bedrock lets development teams build AI applications using AWS-native security and governance controls. Codex on Amazon Bedrock brings OpenAI's leading engineering agent, used by more than five million people every week, into AWS so teams can write, review, debug, and modernize code where they already build and ship.

Why OpenAI on AWS Removes a Major Adoption Barrier

The friction in enterprise AI rarely comes from the models themselves. It comes from procurement cycles, security reviews, billing, and the long road from a promising demo to something approved for production.

OpenAI frames the AWS path as a way to collapse that gap. By making its capabilities available through familiar AWS environments across both Commercial and GovCloud regions, the company says organizations can spend less time clearing operational hurdles and more time building. Teams get OpenAI's frontier capabilities, a known AWS operating model, and a shorter route to deployment in one place.

Amgen CTO Sean Bruich put the stakes plainly. According to OpenAI's announcement, Bruich said GPT-5.5 and frontier models offer "compelling advances in capability, quality, and consistency" in a field where scientific accuracy and decision quality set the standard. For Amgen, having those models on AWS means evaluating them within the responsible AI framework the company already runs.

What Comes Next, Including Daybreak

OpenAI calls this launch the start of a broader effort, not the finish line. The company says it will keep expanding the capabilities offered through AWS so teams can move from evaluation to production with less friction.

One piece of that roadmap is Daybreak, which OpenAI describes as its vision for how software gets built and defended. Daybreak includes cyber models and Codex Security, and OpenAI says it is meant to help defenders spot risk earlier and respond faster by folding secure code review, threat modeling, patch validation, dependency risk analysis, and remediation guidance into the daily development loop. The company has flagged it as a future availability rather than something shipping today.

For security teams, the draw is the same as the broader launch. As specialized tools like Daybreak arrive, AWS gives them a path to adopt those capabilities through the governance and procurement frameworks they already use.

The open question now is timing. OpenAI has not given a public date for when Daybreak or its cyber capabilities will reach general availability on AWS, and that is the detail enterprise buyers will be watching for next. OpenAI's full announcement covers the available models and steps to get started.