Agentic AI Use Cases for Security Soar, but Risks Demand Close Attention

**Executive summary:** This article reviews the growing use of agentic AI in cybersecurity, noting that organizations are increasingly deploying AI agents to automate triage, surface indicators of compromise, accelerate learning curves, and free analysts for higher-value work; industry experts emphasize productivity and support rather than replacement. It also warns of downsides—skill erosion, automation dependency, and the risk that attackers will target AI-driven software—urging careful adoption and human oversight.