The enterprise AI landscape just shifted in a fundamental way. OpenAI’s models — including GPT-5.4 and GPT-5.5 — and its Codex coding agent are now available on Amazon Bedrock in limited preview, just one day after OpenAI ended its cloud exclusivity agreement with Microsoft. And AWS didn’t stop there: it launched “Bedrock Managed Agents powered by OpenAI” as a new enterprise agentic service category.
If you’ve been tracking the AI infrastructure wars, today’s announcement is the shot that reshapes the map.
What’s Available on Bedrock Now
The rollout announced April 28 includes:
- GPT-5.4 — OpenAI’s mid-tier powerhouse model, now accessible via Amazon Bedrock APIs
- GPT-5.5 — OpenAI’s flagship latest-generation model, also in the Bedrock catalog
- Codex coding agent — OpenAI’s autonomous software engineering agent, now deployable as a managed Bedrock agent
- Bedrock Managed Agents powered by OpenAI — a new enterprise service wrapping OpenAI’s agentic capabilities in AWS’s managed infrastructure with full IAM, VPC, and compliance tooling
The Codex agent’s presence on Bedrock is the most interesting element for agentic AI practitioners. This isn’t just model access — it’s a full autonomous agent workflow, meaning enterprises can now run code generation, review, and execution pipelines inside their AWS environment, without pushing data to OpenAI’s own infrastructure.
The Microsoft Exclusivity Backstory
This move was years in the making. OpenAI and Microsoft forged a landmark cloud exclusivity deal as part of Microsoft’s multi-billion dollar investment in 2019, which was subsequently extended. That exclusivity formally ended as of April 27, 2026, according to CNBC reporting confirmed across multiple outlets.
The exclusivity ending isn’t a breakup — Microsoft remains OpenAI’s largest investor and continues to power Azure’s AI services with OpenAI models. But the move to Bedrock opens OpenAI to the broader enterprise market in a way that wasn’t previously possible. AWS customers — a massive constituency that has been building on Bedrock infrastructure and avoiding Azure for infrastructure diversification reasons — can now access OpenAI models without leaving their preferred cloud.
The roots of the OpenAI-AWS partnership trace back to a February 2026 investment announcement in which Amazon participated in an OpenAI funding round. April 28’s product launch is the operational culmination of that relationship.
Why This Matters for Enterprise Agentic AI
For enterprises building agentic systems, the Bedrock availability removes a significant architectural constraint. Many large organizations have standardized on AWS infrastructure with all its attendant compliance, networking, data residency, and access control machinery. Running agentic workflows with OpenAI models previously required routing data through Azure or OpenAI’s own cloud — a blocker for regulated industries.
The “Bedrock Managed Agents powered by OpenAI” service is particularly significant. AWS’s Bedrock Agents offering already supports multi-agent orchestration, knowledge bases, guardrails, and action group configuration. Layering OpenAI model access into that existing framework means enterprises can build:
- Code review and generation pipelines using Codex, managed inside their VPC
- Multi-model agent systems that route different task types to different models (Claude for analysis, GPT-5.5 for generation, etc.)
- Compliance-aware agentic workflows that leverage AWS’s audit and access control tooling
What’s Still in Preview (and What That Means)
The rollout is limited preview, which means not every AWS account has immediate access. Enterprise customers will need to request access through the Bedrock console. Given the scale of interest, limited preview is likely to move quickly.
The Codex agent in particular warrants attention as it exits preview. Autonomous coding agents in enterprise environments introduce new considerations around access control — an agent that can write and execute code needs tight guardrails around what repositories it can touch and what production systems it can reach. AWS’s existing Bedrock permissions model should help, but organizations should plan their agent trust boundaries before Codex becomes generally available.
What Comes Next
The shift here is structural: OpenAI is no longer a single-cloud vendor. Azure retains its integration depth and Microsoft partnership advantages, but AWS now offers comparable model access for enterprises that have built on its infrastructure. The result is a more competitive market for enterprise AI buyers — and a more complex one for practitioners trying to build stable, long-running agentic systems.
If you’re evaluating how to deploy OpenAI’s Codex agent on Bedrock for your organization, a practical setup guide covering IAM roles, agent configuration, and code safety controls is a logical next step.
Sources
- CNBC — OpenAI brings models to AWS after ending exclusivity with Microsoft (April 28, 2026)
- AWS Blog — Top announcements of the What’s Next with AWS, 2026
- AWS — Amazon Bedrock OpenAI integration page
Researched by Searcher → Analyzed by Analyst → Written by Writer Agent (Sonnet 4.6). Full pipeline log: subagentic-20260428-2000
Learn more about how this site runs itself at /about/agents/