Discover Shadow AI Usage

Shadow AI Discovery shows which AI tools are actually being used across your organization — including the unsanctioned (“shadow”) ones people adopt on their own. Open Shadow AI in the sidebar.

The four numbers at the top

• AI tools detected — how many distinct AI tools were seen.
• Not approved tools — detected tools you have not approved.
• Users with unapproved tools — how many people are using something unapproved.
• High-risk tools — tools flagged as high risk.

Two views

Switch between the Tools and Users tabs.

Tools

Every detected tool, with:

• Category — Chatbot, Coding assistant, Image generation, Agent, or Other.
• Policy — Approved or Not approved.
• Detection — how it was seen (for example Endpoint).
• Users — how many people use it.
• Connections — how many times it was contacted.
• Risk — a risk score for the tool.

Users

Every person, with how many AI tools they use and their most-used tool.

Set what is allowed

Change a tool's Policy between Approved and Not approved to reflect what your organization sanctions. Anything left Not approved is your shadow-AI exposure.

Filter and export

Narrow the view by time range, Source, Status (Approved / Not approved), and Category. Use Refresh to reload, or export the current view to CSV or JSON.

Why it matters

Unsanctioned AI tools can train on or leak your data. This page gives you the visibility to sanction the safe tools and act on the risky ones — turning unknown usage into a managed list.