Claude at work: what your employer can see on Team and Enterprise

Checked November 2026. Part of the guide what your employer can see in your AI chats. Facts below link to Anthropic’s Help Center and Claude Platform docs, and the Sources list at the bottom quotes them.

The short answer

On a Claude Team or Enterprise plan, the organization’s Primary Owner manages your work account and all associated data, including the ability to request your conversations through data exports. Enterprise adds a Compliance API that can pull chat data. The audit logs themselves do not carry chat text. Anthropic’s help pages treat Free, Pro, Max and Ultra as consumer plans with separate privacy terms.

Question	What Anthropic’s pages say
Can admins read chat content?	Team and Enterprise Primary Owners can export conversation data. Enterprise Compliance API reads chats, files and projects.
Audit logs	Event records with identifiers; no chat titles or content. Export covers the past 180 days.
Retention	Enterprise: indefinite by default; custom period, minimum 30 days.
Export	Primary Owner data export; Compliance API for Enterprise.
Personal vs work	Work account is governed by the agreement between Anthropic and your organization.
Network tools	Enterprise Tenant Restrictions can block personal accounts on a corporate network.

What admins can see: content versus metadata

Audit logs (metadata). Anthropic says Organization Owners and Primary Owners can export audit logs from Organization settings, and that all audit logs for the organization within the past 180 days will be aggregated. Events cover sign-ins, project and conversation creation, renames and deletes, file uploads, data exports and similar. The page says the title and content of chats and projects are not available to be exported in audit logs, only their unique identifiers. The article we read does not say which plans have audit logs, so check with your admin rather than assume Team has them.

Data exports (content). Anthropic says organization data exports are only available to Team and Enterprise plan Primary Owners and include conversation data and the user data for the organization. Exports do not include files Claude creates in Word, Excel, PowerPoint or PDF format, and do not include messages that were already deleted before the export.

Compliance API (content, Enterprise). Anthropic says the Compliance API lets your organization programmatically pull activity feed events, chat data, and file content across Claude deployments. It is available to Enterprise organizations, and only the Primary Owner can enable it. The API docs say it covers chats, files and projects in claude.ai organizations and sessions in Cowork, Claude Code, Claude in Chrome and other Claude apps, and that Compliance Access Keys can read chat content and also delete chats, files and projects on demand.

Users can export too, if allowed. Owners can turn on a setting that lets members export their own data. It is separate from the Primary Owner’s full-organization export.

How long it is kept

• Enterprise default. Anthropic says by default, data is retained indefinitely unless a custom retention period is set. An Owner or Primary Owner can set a period for chats, projects and artifacts, with a minimum of 30 days.
• How the clock works. Retention is counted from the last observed activity, and opening a chat counts as activity. Chats inside a project follow the project’s retention period, which is indefinite by default, so a chat retention setting alone does not delete them.
• Not covered. Anthropic says custom retention does not apply to standalone Claude Design, Claude Tag, Claude Managed Agents or other features that run Claude Code in the cloud.
• Team plan. Anthropic’s Team and Enterprise deletion page says prompts and conversations are maintained so you can see your history, and you can delete your chats to remove them from your visible history. It says custom retention timelines are an Enterprise option; we did not find one for Team in the pages read.
• Backend retention. Anthropic’s separate page on organizational data says it retains inputs and outputs for up to 2 years, and trust and safety classification scores up to 7 years, if a chat is flagged as violating its usage policy.
• Activity feed. The Compliance API Activity Feed is retained for 6 years and recording is not retroactive: it starts when the API is first enabled.
• After you are removed. Anthropic says remaining members cannot access a removed user’s chats, but the user’s data will still be included in data exports run by the Primary Owner and, on Enterprise, still follows custom retention settings. Re-adding the same email restores previous chats.
• Retention changes are logged. Retention-related actions and changes are tracked in audit logs.

Search, hold and export

The pages we read describe exports and the Compliance API rather than a built-in legal-hold or eDiscovery console. Anthropic does document security and compliance platform integrations built on the Compliance API, and Microsoft documents a Purview connector for Claude Enterprise (marked as a preview) that supports auditing, eDiscovery and retention in Purview. If your employer uses a legal hold on Claude data, it is probably through one of those routes.

Personal account versus work account

• Work account. Anthropic says the use of your Claude for Work plan account is governed by the agreement between Anthropic and your organization, and that its consumer terms and privacy policy do not apply to it.
• Consumer plans. Anthropic’s help pages label Free, Pro, Max and Ultra as consumer products with their own privacy terms, separate from the Team and Enterprise plans.
• Switching. The pages we read do not say whether a consumer account can be claimed by an employer’s domain. Ask your admin if your work email was used on a personal account.

Browser and network monitoring

For Enterprise, Anthropic documents Tenant Restrictions: a company’s network proxy injects an HTTP header into Claude requests, and Anthropic blocks access from any organization not on the allowed list. Anthropic says this is meant to stop corporate-network users from using personal accounts, and its configuration example says TLS inspection is required. In plain terms, a company can configure its own network to see and block Claude traffic from personal accounts.

Anthropic also has a help page on IP allowlisting. We did not find an Anthropic page describing what an employer’s endpoint tools can capture.

What to do

Short checklists, built only from the pages cited on this page. Where a step is plain advice rather than a vendor statement, it has no link.

Other tools

• Microsoft 365 Copilot
• ChatGPT Enterprise, Business and Edu
• Gemini for Google Workspace
• All four compared
• Can my employer see my AI chats (short answer)
• Does your company need an AI policy?

Sources

All pages checked November 2026.

• Anthropic Help Center, Access audit logs: “all audit logs for the organization within the past 180 days will be aggregated.” “title and content of chats and projects are not available to be exported in audit logs (only their unique identifiers will be exported).”
• Anthropic Help Center, Access the Compliance API: “The Compliance API lets your organization programmatically pull activity feed events, chat data, and file content across all your Claude deployments.”
• Claude Platform Docs, Compliance API and Query the Activity Feed: “Activities are queryable within 1 minute of occurring and are retained for 6 years.”
• Anthropic Help Center, Export your organization’s data: “Organization data exports are only available to Team and Enterprise plan Primary Owners. Data exports include conversation data and the user data for your organization.”
• Anthropic Help Center, Configure custom data retention controls for Enterprise plans: “By default, data is retained indefinitely unless a custom retention period is set.” “The minimum retention period is 30 days”.
• Anthropic Help Center, Can you delete data that I sent via Team and Enterprise plans?
• Anthropic Help Center, Who owns and manages the data of my team?: “Your organization’s designated Primary Owner manages your Work account and all associated data.”
• Anthropic Help Center, What happens to a user’s data when they are removed from a Team or Enterprise organization?
• Anthropic Privacy Center, How long do you store my organization’s data?: “We retain inputs and outputs for up to 2 years and trust and safety classification scores for up to 7 years if your chat is flagged by our automated trust and safety systems as violating our Usage Policy.”
• Anthropic Help Center, Enforce network-level access control with Tenant Restrictions: “preventing unauthorized use of personal accounts.”
• Anthropic Help Center, Restrict access to Claude with IP allowlisting and Get started with Claude Compliance API integrations.
• Anthropic Help Center, Move your personal Claude account to a Team or Enterprise organization: “A blue checkmark shows which account you’re currently using.” and “Enterprise admins can claim all existing personal accounts on their organization’s verified domain and move them into the Enterprise workspace.”
• Microsoft Learn, Purview for Anthropic Claude (Enterprise): “The Anthropic Claude connector is currently in preview.”