Skip to main content
Use policy enforcement to control how users interact with AI tools in managed environments.

​
Core policy controls

  • Model and tool access: Allow, monitor, or block specific AI services
  • Content filtering: Define prohibited content categories and actions
  • On-device redaction: Mask configured sensitive fields before data is sent
  • Exceptions: Assign scoped policy overrides for approved roles or groups
1

Start in monitor mode

Observe usage patterns before applying hard-block controls.
2

Enable redaction for sensitive data types

Start with the data classes that matter most for customer risk posture.
3

Apply targeted blocks

Restrict only high-risk actions first, then expand based on evidence from policy events.
4

Review weekly and tune

Update policy using reporting trends and customer feedback.
Avoid broad, immediate blocking across all users on day one. Start with staged enforcement to reduce operational disruption.

Security Agent overview

Reporting overview