Prisma AIRS AI Runtime Security now supports FedRAMP Moderate environments for API Intercept mode. This enhancement provides federal agencies, government contractors, and regulated enterprise organizations with the full spectrum of inline AI threat detection and guardrail capabilities while operating strictly inside a certified, compliance-isolated cloud boundary. By delivering parity with commercial AI security engines within a government-authorized infrastructure, public sector users can safely accelerate the adoption of Large Language Models (LLMs), generative AI apps, and agentic workflows without requiring net-new security audits.
The feature extends complete threat inspection across API deployment models. Supported AI-specific guardrails include real-time detection and mitigation of prompt injections, sensitive data leakage (DLP), toxic content, malicious URLs, SQL CRUD violations, malicious code extraction, contextual grounding failures, and custom topic restrictions. Inspection, management, and log storage occur within a FedRAMP Moderate boundary, except for URL filtering, which runs outside the boundary.
Activating FedRAMP Moderate support is integrated directly into existing administrative workflows within the Customer Support Portal (CSP) and HUB. When you create or update a deployment profile for AI Runtime Security API Intercept, you can toggle an explicit FedRAMP Moderate environment setting. Choosing this option automatically adjusts the required Flex Credit pool allocation with a standard 15% surcharge to cover the operational overhead of the compliance boundary. Once activated, the remaining onboarding, policy configuration, and log streaming steps mirror standard commercial procedures, ensuring a seamless setup for security administrators.
