# Anthropic Compliance

The Anthropic Compliance integration connects Anthropic's Compliance API to import Claude chats from claude.ai web and desktop, alongside Claude usage and cost metrics. Chat activity feeds security review; usage and cost feed the platform's [Observe](/docs/ai-control-plane/observe) and [Costs](/docs/ai-control-plane/observe/costs) views.

## Access requirements

<Callout type="info">
  Viewing connection status on the AI Integrations page requires the `org:read` scope. Connecting, configuring, or deleting the integration requires the `org:admin` scope, held by the default [Admin role](/docs/ai-control-plane/org-admin/roles-and-permissions).
</Callout>

## Create the credential

Anthropic's Compliance API needs two values: an API key scoped to compliance access, and your organization ID.

1. In Claude's [API access settings](https://claude.ai/admin-settings/api-access), create a **Compliance Access API Key**. Grant it these scopes:

   `read:compliance_activities`, `read:compliance_org_data`, `read:compliance_org_settings`, `read:compliance_user_data`, `read:analytics`, and `read:spend_limits`.

   Leave `delete:compliance_user_data` and `write:spend_limits` unchecked, since the integration only reads data. Copy the key; Claude shows it once.

2. Copy your organization ID from Claude's [organization settings](https://claude.ai/admin-settings/organization).

See Anthropic's own [compliance API access guide](https://platform.claude.com/docs/en/manage-claude/compliance-api-access) for the authoritative walkthrough of generating this credential.

## Connect it in the dashboard

Open **Organization settings > AI Integrations**, find the **Anthropic Compliance** provider, and click **Configure**. Enter:

- **Compliance Access API Key**: the key from step 1 above.
- **Anthropic Organization ID**: the `org_...` value from step 2 above. Required.

Save the configuration.

## Verify the connection

The integration doesn't test the credential synchronously on save; the first scheduled poll does. If the key or organization ID is wrong, Anthropic rejects the request and the affected stream shows a rejected-credential status with a **Retry now** action once you've corrected the credential. A stream that keeps failing auto-pauses after three consecutive rejections rather than retrying indefinitely against a bad credential.

Once polling succeeds, three streams populate independently:

- **Compliance activity feed**: Claude chat message events, for security review. Polls roughly every 5 minutes.
- **Claude Chat usage metrics**: token usage. Polls roughly every 4 hours, matching how often Anthropic refreshes its own analytics export.
- **Claude Chat cost metrics**: cost in USD. Same cadence as usage metrics.

Each stream can be paused or resumed on its own from the connection's detail row.