Summary
On September 9, 2026, Zscaler launched Agentic SOC, an AI-first security operations offering that pairs its inline zero trust telemetry and a global decoy mesh network with expert-validated AI agents for triage, root-cause investigation, verdicting, and automated containment. It is generally available globally and uses frontier models from Anthropic and OpenAI.
What changed
Zscaler released Agentic SOC, combining inline zero trust telemetry, a global decoy mesh network, and specialized AI agents built on Anthropic and OpenAI models that triage alerts, investigate root cause, render verdicts, and automate containment. It is available globally at launch.
Why it matters
Applying autonomous agents to the SOC targets alert fatigue and slow response as attackers themselves move at machine speed. Grounding the agents in Zscaler's proprietary inline telemetry is a defensibility play: in agentic security, the data moat, not just the model, is the differentiator.
Evidence excerpt
Zscaler announced Zscaler Agentic SOC, a new approach to security operations built to proactively reduce exposures, scale human expertise and stop AI-driven attacks at machine speed. Available globally today. Combines inline zero trust telemetry, a global decoy mesh network, and expert-validated AI agents.