What happened
Tigera announced GA of Tigera Lynx on June 17, 2026 — a Kubernetes-native control plane that discovers every AI agent in a Kubernetes estate, assigns cryptographic identities, enforces policy on every action, sandboxes agents, audits activity, and detects anomalous behavior — all without changing agent code. Built on Tigera's decade of Calico-based Kubernetes network security.
Why it matters
Addresses the governance gap in Kubernetes-hosted agentic AI: most agent security tools operate at the application layer and require code changes, while Lynx enforces at the network/policy layer with zero agent modification — directly relevant for AI workloads running on K8s.
Applicability
Platform engineering and security teams running AI workloads on Kubernetes should evaluate Lynx for agent identity and policy enforcement; particularly relevant for organizations already using Calico/Tigera for cluster networking.