AI Defense Policies

Create policies using built-in guardrails and templates to ensure data protection, legal compliance, and responsible AI use. These policies are applied to your connected endpoints, helping to secure and govern their interactions.

To create a new policy, click Create Policy on the right top corner of the Policies page. Select from Gateway and API.

To create a new policy:

Note

A left slide-in pane is displayed. This pane has a step-by-step guide for creating a new policy.

Procedure

Step 1

On the New Policy page, under Policy Details, enter a name for the policy.

The Description is optional, you can add details if required.

  1. Click Next.

Step 2

A list of connections is displayed, you can select the checkbox for one or multiple connections to apply the policy. Click Next.

Step 3

Under Security guardrails(prompt injection and code detection), configure security rules.

  1. Move the slider to Enabled for the rule.

  2. You can click the dropdown for Rule directionality to apply this rule to prompts, responses or both.

  3. Click the dropdown for Action to select the action this policy rule will take. You can select: Block or Allow.

  4. Click Next.

Step 4

To protect data and maintain confidentiality, under Privacy guardrails, there are three rules:

  1. Protected health information (PHI):

    1. Move the slider to Enabled.

    2. For Rule directionality, click the dropdown for Directions to select Prompts, Responses or Both.

    3. Click the dropdown for Action to select the action this policy rule will take. You can select: Block or Allow.

  2. PII (Personally Identifiable Information):

    1. Move the slider to Enabled.

    2. Slide to Enable the one or multiple PII entity.

    3. Click the dropdown for Directions to select Prompts, Responses or Both.

    4. Click the dropdown for Action to select the action this policy rule will take. You can select : Block or Allow.

  3. Payment Card Industry (PCI):

    1. Move the slider to Enabled.

    2. Choose the entities for which you want to apply the subcategories and their direction and rule action.

    3. For Rule directionality, click the dropdown for Directions to select Prompts, Responses or Both.

    4. Click the dropdown for Action to select the action this policy rule will take. You can select: Block or Allow.

  4. Click Next.

Step 5

Under Safety guardrails, configure security rules.

  1. Move the slider to Enabled for the rule.

  2. For Rule directionality, click the dropdown for Directions to select Prompts, Responses or Both.

  3. Click the dropdown for Action to select the action this policy rule will take. You can select: Block or Allow.

  4. Click Next.

Step 6

On the Summary page, review your policy details. Click Save to create the policy.

The policies are disabled by default, you would need to enable the policies.

The policy is created and displayed under Policy.