Back Issues/Search Home → Calendar → Archive → Current Issue → Popular →

All issuesVolume 341, Issue 3IT NewsCxO

Inside a CISO's Playbook: What Agentic Security Looks Like in Practice

Cyber Defense Magazine, Tuesday, August 18th, 2026

Cyberhaven's security team replaced human-in-the-loop triage with autonomous AI agents acting as force multipliers.

Facing a roughly 200:1 ratio of engineers to AppSec professionals, Cyberhaven's Office of the CISO shifted from AI-assisted tools to autonomous security agents that reason, collaborate and execute.

Its Cerberus agent runs a multi-model ensemble from Anthropic, OpenAI and Google that performs reachability analysis against the real codebase, then cross-examines findings before a high-reasoning judge model issues a verdict.

The multi-vendor design cut false positives 85% versus raw scanner output. A separate threat modeling agent extends architectural review into the RFC process, turning security from a bottleneck into an orchestration layer.

more →  ·  More from CxO →