Anthropic released a batch of security and admin controls for Claude Enterprise on November 5, headlined by a new security center and the general availability of inference hooks. The update is aimed at security teams that want to apply their existing data loss prevention and monitoring tools to Claude chat, Cowork and Claude Code without managing each product separately.
Inference hooks are the one piece moving to general availability. Prompt hooks send prompts from supported Claude products to an organization's security tool for a check, and pre-tool-call hooks check tool calls before they run. Anthropic says the goal is for DLP and security policies to apply to prompts and tool calls as people work. Hooks are managed in Organization settings.
Reco, one of the vendors that sits behind the hooks, describes the flow: a governed prompt goes to its security server first, which evaluates the transcript, prior tool calls and attachment text, then allows or denies the request before Claude sees it. Reco says the server has a few seconds to return a verdict.
Key facts
- 01Anthropic announced the Claude Enterprise security updates on November 5, 2026.
- 02Inference hooks, which send prompts and proposed tool calls to an organization's security tools, are now generally available.
- 03A beta security center compares chat, Cowork and Claude Code settings against Anthropic's recommended configuration.
- 04A beta parent organization console at admin.claude.com centralizes SSO, SCIM, domain verification, groups and Compliance API keys.
- 05Admins can run their EDR sensor inside the local Cowork virtual machine on macOS and Windows, in detection mode only.
The security center, in beta, is a new page in Organization settings that compares an organization's Claude configuration against Anthropic's recommended settings across chat, Cowork and Claude Code. For each control it shows the current value, the recommended value where one exists and a link to change it, and it flags settings that are new or recently changed. Admins can record attestations, accept risks and export a report for auditors. Anthropic describes the results as guidance, not a certification of compliance with any law or standard.
Larger customers with several Claude Enterprise and Claude Platform organizations get a parent organization console, also in beta, at admin.claude.com. It centralizes single sign-on, SCIM directory sync, domain verification, groups and Compliance API keys, and supports multiple identity providers, so an acquired company can keep its own. Each organization keeps its own members and security controls, and parent roles do not grant access inside an organization by default. Setup requires approval from admins across the organizations and cannot be undone.
Anthropic also added a beta Admin API call that signs a member out of every browser and device without removing them from the organization. It ends sign-in sessions and forces apps such as Claude Code to sign in again, but it does not revoke API keys, cloud sessions already running or scheduled routines.
Other beta controls include a self-serve IP allowlist that admins can configure without going through Anthropic, Claude Desktop settings that can be set without a device management push, the ability to block personal Claude accounts on managed computers, and owner limits on which workspace folders people can use with Claude.
For endpoint security, admins can now run their organization's EDR sensor inside the Cowork virtual machine on a user's computer, reporting to the console their security team already monitors. Claude verifies the vendor's package signature each time the VM starts. The sensors run in detection mode, reporting activity without blocking it, and cover the local VM rather than Cowork sessions in the cloud. The feature 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.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
