Monitor and Govern Claude Enterprise Activity with Lumia Security
Claude is becoming part of everyday enterprise work, helping employees analyze documents, write code, and delegate tasks to AI agents. Security teams need visibility into those workflows: what information is being shared, how Claude is being used, and whether that usage follows company policy.
Lumia Security’s Claude Compliance API integration brings Claude Enterprise activity into Lumia’s AI Usage Control platform. Security and compliance teams can analyze Claude interactions against organization-specific policies, investigate sensitive-data exposure, and review employee and agent activity alongside other enterprise AI usage.
This guide explains what the integration does and how to connect your Claude Enterprise organization to Lumia.
Understand Claude Usage in Your Business Context
Lumia helps enterprises secure and govern how employees and AI agents use AI applications. Its platform analyzes the content, context, and intent of AI interactions, enabling organizations to define policies in natural language around the information and activities that matter to their business.
The Claude integration extends that analysis with data retrieved directly from Anthropic’s Compliance API. This complements Lumia’s existing network-level visibility and controls with an organization-level source of Claude activity.
Security teams can use the integration to:
- Identify sensitive-data exposure: Detect confidential business information, personal data, credentials, and other sensitive content in supported Claude interactions.
- Monitor organization-specific policies: Evaluate usage against policies tailored to your business, such as restrictions on sharing unreleased product plans or customer information.
- Investigate activity with context: Review relevant content and activity together to understand what happened and who was involved.
- Support compliance reviews: Use collected activity and findings as evidence of AI usage oversight.
For example, an organization could define a policy to flag conversations containing confidential acquisition plans. Lumia can evaluate relevant Claude content against that policy and surface findings for security review.
What the Integration Covers
The Claude Compliance API makes Claude Enterprise conversation content, uploaded files, projects, session transcripts, and activity events available to authorized integrations. Session transcripts can include Claude Code and Cowork prompts, responses, and tool-call content.
Coverage depends on the customer’s Claude deployment, the permissions granted to the integration, and the data available through the API. Claude Platform activity feeds do not include model-inference prompts and responses.
The Compliance API connector supports monitoring and investigation after activity occurs. It complements Lumia’s real-time controls; it does not itself block a prompt or stop an agent action before execution.
Configuring the Integration
As a prerequisite, you would need to enable Claude Compliance API for your organization, and provision an API key scoped to Claude Compliance API. Instructions for that are in Anthropic's documentation.
Once you have, simply login to Lumia's portal, navigate to Settings > Integrations and select "Claude Compliance API".
Give a name to the integration and paste in the provisioned API key. It will be stored securely in Lumia's backend and be used to pull in all the Claude interaction data from the Compliance API.
