Safe AI Workbench Developer Docs

Policy Management API

Define and enforce compliance rules for AI interactions across your organization

Configurable Compliance Controls

The Policy Management API enables administrators to define rules that govern how AI systems handle sensitive data. Policies can detect, block, redact, or audit content based on organizational requirements and regulatory obligations.

PHI Detection Policies

Configure how Protected Health Information is detected and handled, with support for blocking, redacting, or allowing with audit.

Hierarchical Scoping

Apply policies at organization, group, or user level with priority-based conflict resolution.

Real-Time Alerts

Receive notifications when policies are triggered, enabling rapid response to potential compliance issues.

Model Controls

Restrict which AI models can be used and configure fallback behavior when primary models are unavailable.

Key Capabilities

  • Create policies with customizable detection rules and enforcement actions
  • Define content filters for specific data types and entity categories
  • Set retention periods for audit logs and conversation history
  • Configure approval workflows for high-risk AI operations
  • Enable or disable policies without deleting them
  • Track policy effectiveness through analytics dashboards

Ready to integrate?

Full API documentation with code samples, authentication details, and interactive examples is available to customers in the product documentation.

Request API Access