Skip to content
Neural Daily

BusinessNews

Claude Enterprise: security center and inference hooks

🇭🇺 Olvasd el magyarul →

Anthropic shipped a Claude Enterprise security pack: a security center (beta), generally available inference hooks, a parent admin console, Admin API sign-out, self-serve IP allowlists and EDR inside the Cowork VM.

· 2 min read

Claude Enterprise security updates Open Graph cover from Anthropic
Image: Anthropic · Press kit · Source

What is GA vs beta

Anthropic made inference hooks generally available for Claude Enterprise: prompt hooks send prompts, and pre-tool-call hooks send tool calls, to your existing DLP or security stack before Claude continues.

The security center is in beta. It compares chat, Cowork and Claude Code settings with Anthropic's recommended configuration on one page, with attestations and a downloadable auditor report. It is guidance, not a certification.

Admin, access, EDR

Centralized management (beta) gives a parent org console at admin.claude.com for SSO, SCIM, domains and Compliance API keys. New access controls include signing a member out via one Admin API call across browsers and devices, a self-serve IP allowlist, and blocking personal or outside Claude accounts in Claude Desktop on managed computers.

An EDR integration (beta) installs your sensor inside the local Cowork/unified Claude VM in detection mode. For earlier Enterprise cost controls, see smart reports and pooled budgets.

Key takeaways

  • Inference hooks are GA for prompt and pre-tool DLP checks.
  • Security center, parent admin and Admin API sign-out are beta.
  • EDR in the local Cowork VM plus Desktop shadow-IT blocks.

    By Mikola Máté Patrik

    Produced with the help of AI, reviewed by an editor. How we make our articles

    Reviewed and edited by the Neural Daily desk. About

    Related articles