search

Viewing, Allowing, and Blocking Your AI BOMs

hashtag
Viewing AI BOMs

You can view all of your AI BOMs or filter them by asset type. For a list of supported prompts, see What Can I Ask UnifAI?

Ask UnifAI to show all of the discovered AI BOMs. Here's an example prompt:

Show all discovered AI BOMs

UnifAI shares a breakdown of AI BOMs found in AI Model, MCP Server, and AI Agent.

For example:

Found 719 AIBOM
Breakdown:
• Component: 458
• Aimodel: 122
• Mcpserver: 98
• Aiagent: 41
Suggestions: Enable suggested policies, View all policies

On the right of the screen:

  • Total Assets: Shows the total number of your assets and provides a detailed breakdown of those secured by a policy and those that are not secured.

  • By Asset Type: Displays AI BOMs by type: AI Agents, AI Models and MCP Servers. For each type, see the number of assets that are secured by a policy and those that are not secured.

  • AI Assets By Type: Displays the details of each AI asset, including the Allowed and Blocked assets.

    • AI Models: Lists the models, version, project the model belongs to, organization it belongs to, and the status. Click on each AI model to see the trust score for each AI model. The trust score is computed across 6 weighted dimensions: Data Governance (25%), Security Controls (25%), Compliance (15%), Auditability (10%), Safety (15%), and Transparency (10%). Each dimension scores 0-100 based on disclosed controls, policies, and certifications. Transparency reflects the ratio of publicly documented items vs. enterprise-only or undisclosed items.

    • MCP Servers: Lists the MCP server, project the server belongs to, organization it belongs to, and the status.

    • AI Agents: Lists the AI Agent, project the AI Agent belongs to, organization it belongs to, and the status.

Expand each section to see the following details:

  • Name: The name of the model.

  • Version: The version number of the model.

  • Project: The component of a product. For example, source repository or micro service.

  • Organization: The organization the project belongs to.

  • Asset Status: Model secured by policy (Allowed) or not secured (Disallowed).

hashtag
Allowing or Blocking Assets

In the Asset Status column, you can use the toggle to set the asset status to Allowed or Blocked. Asset Status indicates whether the asset is secured by a policy. An asset with a status of Allowed is governed by an active policy. An asset with a status of Disallowed is not secured by policy.

Last updated