Agent Risk Manager helps you discover and monitor the AI agents operating across your organization and detect security risks. In this guide, you'll learn how to access Agent Risk Manager, navigate it, connect your first integration, configure your policies, and review your AI activity.
For more detailed articles and topics, see our Agent Risk Manager Overview.
Access Agent Risk Manager
How you access Agent Risk Manager depends on whether you use the classic KnowBe4 console or the KnowBe4 Platform. Follow the steps in the matching subsection.
Classic KnowBe4 Console
To access Agent Risk Manager from the classic KnowBe4 console, follow these steps:
- Log in to your classic KnowBe4 console.
- In the top-left corner of the screen, select the grid icon. A drop-down menu will display.
- In the drop-down menu, select Agent Risk Manager.
KnowBe4 Platform
To access Agent Risk Manager from the KnowBe4 Platform, follow these steps:
- Log in to your KnowBe4 Platform.
- In the sidebar, select Agent Risk Manager. A pop-up menu will display.
- Select the tab you want to navigate to, such as Dashboard.
Navigate Agent Risk Manager
To move between pages in Agent Risk Manager, use the sidebar. The sidebar includes the following sections:
- The Dashboard contains a summary of AI activity across your organization.
- The Discovery section contains your organization's AI Activity Log and Inventory.
- The Policies section contains your organization's AI Policies, AI Tool Permissions, and Data Detection Policies.
- The Setup section contains Integrations, where you can connect and manage your integrations.
- The Users section contains User Activity, where you can review the activity of your users.
Connect Your Integrations
Before Agent Risk Manager can discover agents or surface detections, you'll need to connect at least one integration in Setup > Integrations. Integrations connect Agent Risk Manager to your AI vendors so that Agent Risk Manager can receive and analyze agent activity.
Agent Risk Manager currently supports the following integrations:
- Anthropic Claude Compliance
- Google Gemini Enterprise
- KnowBe4 Browser Extension (Intune)
- Microsoft Copilot
- OpenAI ChatGPT Compliance
Setup steps vary by integration. To add your first integration, see our integrations guide.
Configure Your Policies
The browser extension can warn your users when they use AI tools, block them from accessing certain AI tools entirely, or redirect them to an approved alternative. To control which AI tools your users can access, add each tool to the AI Tool Registry.
To add an AI tool, follow these steps:
- In the sidebar, select Policies > AI Tool Permissions.
- In the AI Tool Registry card, select Settings > Add AI Tool.
- Complete the following fields:
- Tool name: The name of the AI tool.
- Vendor: The name of the AI tool vendor. This field is optional.
- Domain pattern: The hostname of the AI tool, without https:// and without a path.
- Tier: The access level for the AI tool. Select Approved, Warned, or Blocked. New tools default to Blocked.
- Block message or Warning message: The message your users see when they open the AI tool. This field displays for Blocked and Warned tools.
- Select Add tool.
Your changes save automatically. For more information, see our AI Tool Permissions Guide.
Review Your AI Activity
To review your AI activity, navigate to the sidebar and select Dashboard. Use the date range picker to set the period for most of the figures on the page. For more information, see our Dashboard guide.
To review events that Agent Risk Manager has flagged as potentially risky, navigate to the sidebar and select Discovery > AI Activity Log. To narrow the table, use the date range picker, the search box, or the filters on each column. To view the details of an event, select its row. For more information, see our AI Activity Log Guide.
Agent Risk Manager groups detections into categories, such as Excessive Agency, Prompt Injection, Sensitive Information, and Unbounded Consumption. For definitions of each category, see our Events and Detection Categories guide.