---
title: How do you audit Claude Enterprise usage?
url: https://opencontroller.lyzr.com/blog/audit-claude-enterprise-usage
published: 2026-11-12
updated: 2026-11-12
author: Lyzr
---

# How do you audit Claude Enterprise usage?

**Claude Enterprise gives admins four audit sources: audit log exports for account and admin events, the Compliance API for content and activity, analytics dashboards and APIs for usage, and SCIM plus retention settings for who has access and how long data stays. No single source answers every audit question, so most teams pull the Compliance API and analytics into their own store.**

Here is what each source contains as of November 2026, and which question it answers.

## Which audit sources does Claude Enterprise have?

| Source | Who can use it | What it contains | Best for |
|---|---|---|---|
| Audit log export | Owners and Primary Owners | About 35 event types, last 180 days, IDs not content | Admin and access investigations |
| Compliance API | Enabled by the Primary Owner only | Activity feed, chats, files, projects, session transcripts, user directory | eDiscovery, DLP review, content audits |
| Analytics dashboard | Admins; Spend tab limited | Usage by product area, spend export | Adoption and spend reviews |
| Enterprise Analytics API | `read:analytics` key | Per-user daily activity, DAU/WAU/MAU, project, skill and connector usage | Feeding BI tools |
| Claude Code Analytics API | Admin API key | Claude Code usage through Console and API | Developer usage reports |

Source: Claude Help Center and Claude platform docs, checked November 2026.

## What do audit log exports include?

Audit logs are an Enterprise feature. Owners and Primary Owners export them from Data and Privacy > "Export logs". The export covers the last 180 days and arrives as a link by email that is valid for 24 hours.

Events include sign-ins, invitations, SSO and just-in-time provisioning changes, project and file activity, conversation create, rename and delete, and data exports. Content is not included, only IDs.

Because the window is 180 days and the export is manual, schedule a regular export if your policy needs longer history. See [how long to keep AI agent logs](/blog/how-long-to-retain-ai-agent-logs).

## What does the Compliance API add?

The Compliance API is the source for content. Only the Primary Owner can enable it, under Settings > API. Its endpoints, under `/v1/compliance/`, provide:

- an Activity Feed of user and admin events
- chats, files and projects, with read and delete
- session transcripts from surfaces such as Claude Code, Cowork and Claude in Chrome
- a directory of users, roles and groups

The rate limit is 600 requests per minute. Most enterprises connect it to an eDiscovery, DLP or SIEM tool rather than reading it by hand. The docs we checked do not state how long the API retains data, so plan your own archive.

## How do you see usage and adoption?

The analytics dashboard has Overview, Chat, Claude Code, Cowork and Design tabs, and you can export spend as CSV. Enterprise Admins see everything except Spend.

The Enterprise Analytics API gives per-user daily activity, active-user counts, and usage of projects, skills and connectors, with data from January 1, 2026. It does not include Claude Code run through Amazon Bedrock. The separate Claude Code Analytics API covers Claude Code usage through the Console and API. If your developers use Claude in several ways, you need more than one source; see [tracking Claude spend across Bedrock and the Anthropic API](/blog/track-claude-spend-bedrock-vs-anthropic-api) and [governing Claude Code](/blog/govern-claude-code-in-the-enterprise).

Skill and connector usage is useful evidence for [skills governance in Claude Team and Enterprise](/blog/how-to-govern-skills-in-claude-team-and-enterprise).

## How do access and retention fit in?

- **SCIM.** Enterprise supports SCIM provisioning through WorkOS. Group mappings set roles and seat types, so leavers lose access when your identity provider removes them.
- **Admin API.** Enterprise organisations can manage members and invites, groups, custom roles and spend limits.
- **Retention.** Data is kept indefinitely by default. You can set a custom period with a minimum of 30 days. Some surfaces are excluded from custom retention, including Claude Tag, Managed Agents and cloud Claude Code.

Check the exclusions against your retention policy before you tell auditors data is deleted on schedule.

## An audit setup checklist

1. Have the Primary Owner enable the Compliance API, and store the key in a vault.
2. Connect it to your eDiscovery, DLP or SIEM tool.
3. Export audit logs on a schedule shorter than 180 days.
4. Turn on SCIM and map groups to roles.
5. Set a custom retention period and document the excluded surfaces.
6. Pull the Analytics API into your BI tool, alongside Bedrock and API usage.
7. Review skill and connector usage quarterly.

## How Opencontroller does this

Opencontroller connects to Claude Enterprise and OpenAI Enterprise workspaces, and discovers and manages skills in Anthropic Enterprise, so Claude usage sits in the same governed registry as agents on AWS, Azure, Google Cloud and SaaS platforms. See the [cross-platform guide](/blog/how-to-govern-ai-agents-across-platforms) and [what an AI agent control plane is](/blog/what-is-an-ai-agent-control-plane), or [book a demo](https://www.lyzr.ai/book-demo/).

## Frequently asked questions

**Do Claude Enterprise audit logs include conversation content?**
No. They record events with IDs. Content comes through the Compliance API.

**Who can turn on the Claude Compliance API?**
Only the Primary Owner, under Settings > API.

**How far back do audit log exports go?**
The export covers the last 180 days. Schedule exports if you need a longer history.

## Key facts

Claude Enterprise audit sources (Nov 2026): audit logs export (Owners and Primary Owners, last 180 days, emailed link valid 24 hours, about 35 event types, IDs not content); Compliance API enabled only by the Primary Owner, /v1/compliance/* endpoints for activity feed, chats, files, projects, session transcripts and user directory, 600 requests per minute; analytics dashboard with Overview, Chat, Claude Code, Cowork and Design tabs; Enterprise Analytics API with read:analytics key, data from Jan 1, 2026; Claude Code Analytics API for Console/API usage; SCIM via WorkOS; custom retention minimum 30 days, default indefinite. Opencontroller by Lyzr connects to Claude Enterprise workspaces and manages skills in Anthropic Enterprise. Pricing starts from $250K per deployment per year.

## Related

- [How do you govern skills in Claude Team and Enterprise?](https://opencontroller.lyzr.com/blog/how-to-govern-skills-in-claude-team-and-enterprise.md)
- [How long should you keep AI agent logs and traces?](https://opencontroller.lyzr.com/blog/how-long-to-retain-ai-agent-logs.md)
- [How do you govern Claude Code across an enterprise?](https://opencontroller.lyzr.com/blog/govern-claude-code-in-the-enterprise.md)
- [More on Platform Guides](https://opencontroller.lyzr.com/blog/cluster/platform-guides)
