Governing Claude: Introducing the Linx + Anthropic Integration

Over the past few months, we have spoken with many security and IT leaders who are in the same position: their organizations are deploying Claude broadly across teams, but they have no idea who actually has access to it, what level of privilege those users hold, or what's happening with the API keys their developers have created. They're governing other application in their stack with rigor, but treating Claude as an exception.

That disconnect has always bothered us. Not only because Claude is unique, but because the same identity logic that protects every other application should apply here as well.

When Anthropic launched the Claude Compliance API, we ceased that opportunity. We're genuinely proud to be one of the integration partners building on it. Anthropic is setting the bar for what responsible enterprise AI deployment looks like, and the Compliance API is a direct expression of that: giving security and IT teams the programmatic access they need to govern Claude the same way they govern everything else that matters. Building on that foundation is something we are excited about.

Here's what we built.

The Linx x Anthropic Integration in Action

Setting it up is simple. Follow the step-by-step integration guide and connect in just 60 seconds. Linx begins pulling data automatically.

Once connected, Claude shows up in Linx as a fully governed application, the same way Salesforce, Okta, GitHub, or any other app in your estate does. From that point forward, everything Linx does for your other applications, it now does for Claude.

What You Can See and Do

Account, Role, and Permission Visibility

The first thing security teams want to know is: who has access, and what can they do? With this integration, Linx gives you a complete view of every account provisioned in Claude, the roles assigned to each user, and the groups they belong to.

You can easily surface administrative roles and identify who holds elevated privileges, and drill into the specific permissions attached to each role. For compliance and least-privilege reviews, this alone is a significant step forward.

AI-Powered Access Certifications

The AI-powered access certifications take the heavy lifting off identity and compliance teams, meaning faster turnaround, fewer errors, and a smoother process end to end.

The new integration allows you to easily review and certify access in Claude, using Linx’s AI-powered access review campaigns. The campaign surfaces every user and their entitlement type, along with AI context and recommendations on whether that access should be approved or denied, powered by Anthropic's own models.

So, you're combining Linx's identity governance context with Claude's intelligence to make faster, smarter access decisions. The reviewer doesn't have to make every call from scratch. They have a recommendation in front of them and the data to act on it.

For teams running quarterly or annual access reviews, this changes the economics of the process considerably. What used to require manual triage across hundreds of accounts can now move in a fraction of the time.

API Key Visibility

One of the less visible but high-stakes risks in any AI deployment is unmanaged API keys. Developers create them, projects get deprecated, people leave the company, and the keys keep existing. Then something goes wrong.

Linx allows identity teams to closely monitor API keys created within Claude: when it was created, associated metadata, and enough context for a security team to evaluate whether it should still exist.

This is the kind of visibility that used to require a manual audit. Now it's continuous.

Project and Resource Governance

Finally, with this integration Linx provides the tools for real governance over projects and shared resources within Claude, with an access graph for each project that shows exactly how access is provisioned.

You see which roles exist, which users hold them, and how access flows through the system. For organizations with multiple teams working across different Claude projects, this provides the resource-level clarity that broad account views can't.

Why This Matters

Security teams have spent years building strong identity governance programs: access reviews, offboarding workflows, least-privilege enforcement, audit trails. Claude is a critical app, and should be governed as such. With this integration, it is.

What makes this feel right to us as product builders is that we didn't create a separate Claude governance tool. We extended the identity fabric that our customers already rely on. Claude is governed in Linx as part of a whole, not as an exception. This means it benefits from everything Linx already does, and security teams don't have to learn a new workflow or platform to govern it.

We're proud to of this integration and to be shaping the future of AI security and governance. Being part of the ecosystem that makes Claude governable at scale is something the Linx team doesn't take lightly.

If you’re new to Linx, schedule a demo to experience truly unified identity governance.