Palo Alto Prisma AIRS Lab
Solution overview
Palo Alto's Prisma AIRS (AI Runtime Security) is a platform built specifically to secure AI applications from development all the way through runtime. AIRS gives you AI model scanning to catch vulnerabilities before they hit production, real-time threat detection on inference traffic, prompt injection and jailbreak protection, sensitive data discovery to make sure your AI pipelines aren't leaking things they shouldn't, and full visibility into how AI APIs and third-party models are being used across your environment.
In this lab, you will test some features of Prisma AIRS in action, blocking messages send to an AI model instantly. You will get to try several different attacks and prompt injections. You will see AIRS block not just prompts but also responses for the local AI model. Then, you will log into Strata Cloud Manager where the AIRS configuration is stored, look at how AIRS is deployed, how the security policies are configured, and what logs are generated when a malicious prompt or attack is detected.