Gurpreet Kaur Khalsa

Today, we are excited to introduce Agent Validation as a new evaluation capability in AI Defense: Explorer Edition, the free self-service version of Cisco AI Defense, that is built specifically for agentic AI systems. Agent Validation builds on the agentic security enhancements to Cisco AI Defense announced at Cisco Live, which introduced adaptive red teaming, Policy Studio guardrails, and supply chain discovery for agents.

Cisco’s Integrated AI Security and Safety Framework and our recent work on defining taxonomy constitutions focused on defining and detecting common risks shared among enterprises when deploying AI. However, while most enterprises share a lot of the common risk categories, they are also diverse, and it is impossible to develop a complete taxonomy that would fully cover all customer specific cases.

AI red teaming is easier to understand when you run it yourself AI security can sound abstract until you point a scanner at a real endpoint and watch what happens. A model may answer normal user prompts perfectly well, but still behave differently when a conversation becomes adversarial. A support assistant may follow its public instructions, but still have hidden rules that should never be exposed.