Cloudflare CASB Adds Claude Compliance API Integration for AI Governance
Enable the Claude Compliance API integration in Cloudflare Zero Trust to surface AI compliance findings in the dashboard and automatically create Gateway policies.
Enable the Claude Compliance API integration in the Cloudflare dashboard: go to Zero Trust > Integrations > Cloud & SaaS, add Anthropic, enter key, configure DLP profiles, and monitor findings.
Summary
Cloudflare has extended its Cloudflare Access Security Broker (CASB) to support Anthropic’s Claude Compliance API, enabling security teams to monitor Claude usage directly in the Cloudflare dashboard without endpoint agents. The integration surfaces security findings for projects, project attachments, chat files, chat messages, and artifacts, all of which are scanned for DLP policy violations and displayed with severity levels in the same interface used for Microsoft 365, Google Workspace, and Salesforce. Findings can be automatically turned into Cloudflare Gateway policies, allowing administrators to block uploads, restrict access, or limit functionality for specific users within minutes. The feature works for both Claude Enterprise and Claude Platform, retrieving conversation content via read‑only endpoints and supporting future Activity Feed integration. To enable the integration, users must have a Claude Enterprise account, request API access, and add the Anthropic integration in Cloudflare Zero Trust under Integrations > Cloud & SaaS, then configure DLP profiles if desired. The new CASB coverage is part of Cloudflare’s unified AI governance strategy, which also includes AI Gateway, Secure Web Gateway, and Access with MCP server portals.
The integration is designed to give enterprises full visibility into AI workflows, from API calls to data at rest, and to provide a single platform for monitoring, triaging, and remediating AI‑related security risks. By leveraging the Claude Compliance API, Cloudflare CASB eliminates the need for inline traffic inspection or endpoint agents, making it easier for organizations to adopt AI tools safely at scale.
Key changes
- Integration with Claude Compliance API for real‑time AI usage monitoring
- Detection of projects shared across the organization or subsets of users
- Detection of project attachments violating DLP policies
- Detection of chat files and messages that contain sensitive data
- Detection of provider‑generated artifacts that breach DLP rules
- Findings surfaced in Cloudflare dashboard with severity levels
- Automatic conversion of findings into Cloudflare Gateway policies
- Supports both Claude Enterprise and Claude Platform with read‑only endpoints