On Nov 5, 2026, Anthropic documented new Claude Enterprise security and admin features. Security Center (beta) compares an organization’s settings with Anthropic’s recommended controls across chat, Cowork, and Claude Code; it is guidance, not a compliance certification. Inference hooks are generally available: prompt hooks and pre-tool-call hooks send prompts or tool calls to the customer’s security or DLP (data-loss-prevention) service for allow/deny before the model runs or a tool executes. Also listed: parent-level admin console, Admin API sign-out, self-serve IP allowlists, blocking personal Claude accounts on managed desktops, and EDR sensors inside local VMs in detection mode (beta).
Claude Enterprise: Security Center in beta, inference hooks generally available
Quick Facts
- Published
- Updated
- Source
- Claude by Anthropic
- Category
- Products & Apps
Why it matters
企业能不能把 Claude 接到正式业务,常常卡在「提示词和工具调用能不能按公司规章先审一遍」。
Related entities
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: Claude Enterprise: Security Center in beta, inference hooks generally available. November 5, 2026. https://new.nat6.net/en/news/custom-799876eb927042f7/