Genie Generate a free chatbot for your company website Try it
← Back to Blog

AWS is moving Bedrock Agents to Classic — what teams should do now

Editorial image for AWS is moving Bedrock Agents to Classic — what teams should do now about Developer Tools.

Key Takeaways

  • AWS says Amazon Bedrock Agents is now Amazon Bedrock Agents Classic and will stop accepting new customers on July 30, 2026.
  • Existing customers can continue using Bedrock Agents Classic, and AWS says Bedrock models, Knowledge Bases, and Guardrails are not affected.
  • AWS recommends migrating Bedrock Agents Classic workloads to Amazon Bedrock AgentCore.
  • The larger shift is from an older managed-agent surface toward a more modular agent runtime and tooling platform.
BLOOMIE
POWERED BY NEROVA

Produced by Bloomie for Nerova AI using automated editorial checks. Sources used for factual claims are listed below.

AWS has put a real deadline on one of its early agent-building services. In its June 30, 2026 service availability update, AWS said Amazon Bedrock Agents is now Amazon Bedrock Agents Classic and will no longer be open to new customers starting July 30, 2026.

The key point is that this is not just a naming cleanup. AWS is drawing a line between its older Bedrock agent service and AgentCore, the newer platform it now recommends for teams that want similar capabilities going forward.

What changed on June 30, and what did not

AWS says existing Bedrock Agents Classic customers can continue using the service as normal. Amazon Bedrock itself is not being retired, and AWS says existing Bedrock models, Knowledge Bases, and Guardrails are not affected.

That distinction matters because some headlines will make this sound bigger or scarier than it is. AWS is not shutting down Bedrock. It is putting the original Bedrock Agents service into maintenance mode for new customer acquisition and signaling that its future agent-building path now runs through AgentCore.

Why AWS is steering builders to AgentCore

The migration guide is explicit: AWS recommends moving Bedrock Agents Classic workloads to Amazon Bedrock AgentCore. That recommendation makes sense when you look at how AWS is positioning AgentCore.

According to AWS documentation, AgentCore Runtime is a secure, serverless, purpose-built environment for deploying and running AI agents or tools. AWS also describes it as framework-agnostic, able to work with agent frameworks such as LangGraph, Strands, and CrewAI, while supporting models from Amazon Bedrock, Anthropic Claude, Google Gemini, and OpenAI.

That is the bigger product signal behind this news. AWS appears to be moving from an earlier, more opinionated managed-agent surface toward a broader runtime and tooling layer that can host more kinds of agents, more frameworks, and more model mixes.

Who actually needs to act before July 30

If your team is not yet a Bedrock Agents customer but still planned to use the original service, the deadline is straightforward: AWS says you need to sign up before July 30, 2026. After that, the Classic service will no longer be open to new customers.

If you are already using Bedrock Agents Classic, the issue is less urgent but more strategic. You can keep operating, but you should treat this as a migration planning window rather than a reason to stay put indefinitely. The maintenance-mode guide makes clear that AWS wants future development to land on AgentCore.

For most teams, the practical work is not just moving prompts. It is reviewing orchestration logic, tool connections, action flows, runtime hosting assumptions, and how your monitoring and security model should change when you move to a more flexible agent platform.

The strategic takeaway for AI builders

The interesting part of this announcement is what it says about the agent stack. In 2024 and 2025, a lot of platform vendors packaged agents as single managed features. In 2026, the market is shifting toward more modular layers: runtime, tool connectivity, observability, identity, and web or data access that can work across frameworks and model providers.

AWS is effectively telling developers that the old way of building agents inside Bedrock was useful, but not where it wants the platform to center going forward. AgentCore looks like the new control layer.

If you are building production AI agents now, the right response is not panic. It is architecture discipline. Confirm whether you are on Classic, decide whether you need short-term continuity or a cleaner rebuild, and use this deadline to simplify the parts of your stack that were too tightly coupled to one older orchestration surface.

That is why this news matters beyond AWS alone. It is another sign that enterprise agent platforms are maturing from feature checklists into more opinionated infrastructure bets.

Nerova context

Custom AI agents for business operations

Nerova builds custom AI agents for business operations. Companies use Nerova when they need AI support for customer intake, support, sales follow-up, research, website audits, internal handoffs, and workflow automation.

Nerova can help turn websites, business context, and operational workflows into practical AI systems: website chatbots, single-purpose agents, AI teams, audits, and automation workflows built around a clear business outcome.

Plan the right migration path for your AI agent stack

Use Scope to map which agent workflows should be rebuilt, which can be migrated, and where governance or tooling changes are needed before you scale.

Run an agent platform audit
Ask Bloomie about this article