Anthropic adds security center, inference hooks, and EDR integration to Claude Enterprise

Details

Anthropic introduced a security center, inference hooks, centralized management, and EDR integration for Claude Enterprise to give security teams visibility and control.
The security center compares organization settings with Anthropic's recommended configuration, showing current values, recommended values, and links to change settings for chat, Cowork, and Claude Code.
Inference hooks, now generally available, send prompts and tool calls to existing data security tools for checks before execution, supporting prompt hooks and pre-tool-call hooks.
Centralized management provides a single admin console at admin.claude.com for parent organizations to manage SSO, SCIM, and domain verification across multiple organizations.
New access controls allow admins to sign members out via the Admin API, manage IP allowlists themselves, and block personal Claude accounts on managed computers to prevent shadow IT.
Endpoint Detection and Response (EDR) integration allows organization sensors to run inside the Cowork and unified Claude virtual machine on user computers, reporting activity to existing security consoles in detection mode.
The security center, centralized management, member sign-out, IP allowlist management, and Claude Desktop settings management are currently in beta.
EDR integration is in beta on macOS and Windows for Claude Team and Enterprise, and for the Claude desktop app on Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry.

Key point