Viewing and managing AI Deployment details

After you connect IBM Guardium AI Security to a cloud account, Guardium AI Security investigates, analyzes, and classifies all the AI deployments in that cloud account.

AI Security This content is available through the Guardium AI Security module.

In the main menu, click Configurations>Connections>AI accounts to see the Connections page. The Connections page lists all the AI deployments in the connected AI service providers in a table that provides you with the following information:

  • Environment of the AI deployment's cloud account
  • Region of the AI deployment's cloud account

AI deployment details

Select an AI deployment in the table of the AI deployments page to get a more detailed view of the AI deployment's environment and attack surface. You also see insights into the data, model, and application used by the AI deployment.

The first part of this page provides the following information about the AI deployment:

  • Subscription or Account name of the AI deployment
  • Service type of the AI deployment
  • Region of the AI deployment's cloud account
  • Model name of the AI deployment
  • Model publisher of the AI deployment
  • Model type of the AI deployment
  • Environment type of the AI deployment

The second part of this page provides information about the AI deployment's attack surface. It shows the relationship and data flow and between the vulnerable data, models, and applications that the AI deployment uses. You can find the following information about the AI deployment's attack surface:

  • Types of data in the AI deployment that is exposed to vulnerabilities
  • Models in the AI deployment that is exposed to vulnerabilities
  • Applications that use the vulnerable models or data
  • Flow of data between the vulnerable data, bodes, and applications

The third part of this page provides more details about the AI deployment's vulnerable attack surface in the following tabs:

  • Data
  • Model
  • Application