{"pageModel":{"attributes":{"id":"","name":"120077.dita","viewName":"DitaDetail"},"elements":{"ditaContent":{"name":"DITAContent","value":"
\r\n

Add an AI Guardrails Rule to the Data Loss Prevention Policy

\r\n
\r\n
\r\n

AI guardrails rules operate within the data loss prevention policy to monitor prompts or\r\n response content exchanged with AI application, including uploaded files. The rules define\r\n what types of data to monitor, and whether content should be blocked or only monitored.

\r\n

An AI Guardrails rule must have at least one of the following two criteria defined:

\r\n
    \r\n
  • Data Classifications —\r\n Include files that match data classification of your own making or a built-in data\r\n classification provided by Secure Access.
    • \r\n
  • File Size — Include files that meet\r\n custom size criteria.
    • \r\n
\r\n
\r\n
\r\n
\r\n

Before you begin

\r\n
\r\n
    \r\n
  • HTTPS Inspection must be enabled either on the default Access policy, or at least one access policy rule where the same identity is configured in an AI guardrails rule. For example, if the resource \"Network A\" is configured for an AI guardrails rule, it must also be configured for an access policy rule where HTTPS is enabled. For more information, see Manage Traffic Decryption and Manage Security Profiles.
    • \r\n
  • Full Admin user role. For more information, see Manage Accounts.
    • \r\n
\r\n
\r\n
\r\n

\r\n

Procedure

\r\n\r\n
\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n
1\r\n

Create the AI Guardrails rule and establish general settings.

\r\n
2\r\n

Select Data Classifications to apply to the rule.

\r\n
3\r\n

Make Files Control selections for the rule\r\n

\r\n
4\r\n

Select Identities to apply to the rule.

\r\n
5\r\n

Select Destinations to include for the rule.

\r\n
6\r\n

Select an Action to be performed when data matches the rule.

\r\n
7\r\n

For any destination you can enable and configure an email notification to be sent to users when a violation occurs.

\r\n
\r\n
\r\n
\r\n","ditaVal":"","format":"html"},"bookTitle":{"value":""},"shortDescription":{"value":""}}},"parameters":{"appId":"SecureAccess","topicAlias":"add-an-ai-guardrail-rule"}}