Skip to main content

Prisma AIRS AI Model Security API Reference

Prisma AIRS AI Model Security is a product that allows customers to scan any top of AI model using custom security configurations known as security groups. Model Security’s public API is documented to enable customers to submit scans, retrieve results, and configure their security groups programmatically from their CI environments or other custom workloads.

Quick Start​

  • Base URL: All interactions with the Prisma AIRS AI Model Security APIs must be directed at the following base URL to be routed appropriately. The base URL for AI Model Security is:https://api.sase.paloaltonetworks.com/aims

  • Authentication: All Model Security APIs require OAuth2 bearer tokens for proper authentication. The format of the bearer token follows the standard authorization header: Authorization: Bearer YOUR_TOKEN_HERE. For instructions on how to obtain an OAuth 2.0 token, see the Access Tokens section.

API Architecture​

The API is divided into two distinct planes:

  1. Management Plane: Handles interactions with security groups and security rules.
  2. Data Plane: Handles submitting scans and retrieving scan results (rule evaluations, violations, and per-file results).

Pagination​

List endpoints use offset-based pagination:

  • limit: Number of items per page.
  • skip: Number of items to skip.