Zscaler launches Agentic SOC for 3,832-alert days

Zscaler launched Agentic SOC, an AI security-operations platform aimed at the 3,832 alerts per day that the company says security teams receive across 83 tools, according to its product materials. It groups signals, adds identity, device, and exposure context, and offers analyst-visible recommendations or automated containment; Zscaler says 67% of SOC analysts worry about missing a relevant event and average investigation takes 70 minutes, citing Vectra and IBM. The operator implication is not “replace the SIEM”: pair agentic triage with human approval first, then measure false positives and containment time against the AI-security debt visible in Palo Alto’s results.