> ## Documentation Index
> Fetch the complete documentation index at: https://docs.twenty.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Permissions & Access Control

> Control what AI agents can access and modify in your workspace.

## Overview

AI agents respect your existing permission structure. This is particularly important for teams who want to control exactly what automated AI processes can access or modify in their workspace.

## Assign a Role to an AI Agent

1. Go to **Settings → Members → Roles**
2. Click on the role you want to assign
3. Open the **Assignment** tab
4. Under **AI Agents**, click **+ Assign to AI agent**
5. Select the AI agent from the list
6. Confirm the assignment

## Why Assign Roles to AI Agents?

| Benefit          | Description                                       |
| ---------------- | ------------------------------------------------- |
| **Security**     | Limit what data AI agents can access or modify    |
| **Compliance**   | Ensure AI only processes the data it needs        |
| **Control**      | Prevent unintended actions from AI automations    |
| **Auditability** | Track which actions were performed by which agent |

<Note>
  For AI agents running within workflows, role assignment ensures the agent cannot access or modify data outside its intended scope—even if the workflow has broader permissions.
</Note>

## Related

* [Permissions](/user-guide/permissions-access/capabilities/permissions) — detailed information on creating and managing roles
* [AI Agents](/user-guide/ai/capabilities/ai-agents) — AI capabilities in workflows