Anthropic's Claude Platform goes live in AWS Console for direct sign-ups
- Anthropic and AWS launched Claude Platform on AWS in general availability, letting customers sign up inside the AWS Console and use Claude with AWS accounts. - The key twist is operational: Anthropic, not AWS, runs the platform, while AWS handles IAM, billing, and commitment retirement through native AWS endpoints. - This shifts AI competition toward procurement, identity, and managed agent infrastructure — not just whose model benchmark chart looks best.
AI model access is turning into an enterprise plumbing fight. The news here is not just that Claude shows up in one more place. It’s that Anthropic and AWS have made Claude’s own platform available directly inside the AWS Console, so companies can buy and use the native Claude experience with their AWS accounts instead of setting up a separate vendor relationship. That sounds small. It isn’t. It changes who owns the customer touchpoint, the billing path, and the control layer around the model. ### What actually launched? Claude Platform on AWS is now generally available. AWS says it is the first cloud provider to offer Anthropic’s native Claude Platform experience this way, including APIs, console access, and early-access beta features through an existing AWS account. Anthropic also published docs and release notes tying Claude Managed Agents into this setup, so the launch is not just about inference calls — it includes agent tooling and runtime features too. (aws.amazon.com) ### So is this just Bedrock? No — and that distinction is the whole story. On Amazon Bedrock, AWS operates the inference stack. On Claude Platform on AWS, Anthropic operates the platform, but customers authenticate and pay through AWS. Anthropic’s own docs frame Bedrock as the legacy AWS integration for Claude, while this new route is the Anthropic-operated alternative with native platform features and often faster feature access. (aws.amazon.com) ### Why would customers care? Because enterprise adoption often dies in procurement, not in evals. This setup removes separate Anthropic credentials, contracts, and billing relationships. AWS IAM policies control access, AWS billing stays consolidated, and AWS commitments can be retired against usage. For a big company, that means less vendor sprawl, one audit trail, and fewer internal approvals before developers can start building. (platform.claude.com) ### What’s the catch? The clean buying experience does not mean the service sits fully inside AWS in the usual sense. AWS explicitly says Claude Platform on AWS is operated by Anthropic and that customer data is processed outside the AWS security boundary. So the sales motion feels AWS-native, but the operating model is different from a standard AWS-managed service. That matters for compliance teams, security reviews, and anyone who assumed “in the console” automatically meant “fully inside AWS.” (aws.amazon.com) ### Where do managed agents fit in? This is where the launch gets more strategic. Claude Managed Agents gives developers a hosted agent harness — basically the loop, runtime, tool execution, file handling, web access, code execution, and session management that teams otherwise have to build themselves. Anthropic says it is also available on Claude Platform on AWS, with some feature differences. In other words, AWS is no longer just a place to call a model. It is becoming a place to procure an entire agent stack run by the model vendor. (aws.amazon.com) ### Why does that matter now? Because model quality is getting harder to use as the only wedge. The more similar frontier models feel in everyday enterprise use, the more the fight shifts to convenience — identity, billing, governance, feature velocity, and whether the vendor gives you a usable execution layer instead of just an API endpoint. Anthropic is basically saying: keep your AWS controls, but get our native platform and our agent runtime. (platform.claude.com) ### Does this weaken Bedrock? Not exactly. Bedrock still exists, and Anthropic says Claude remains available there. But it does create a two-track AWS story: Bedrock for the AWS-operated path, Claude Platform on AWS for the Anthropic-operated path. That gives customers more choice, but it also makes clear that the cloud console is becoming a distribution channel for model companies, not just a hosting layer. (aws.amazon.com) ### Bottom line? This launch is really about control of the enterprise wrapper around AI. The model still matters. But the winning product increasingly looks like model plus identity plus billing plus agent runtime. Anthropic just moved that whole bundle into AWS’s front door. (aws.amazon.com) (platform.claude.com)
