In regulated industries, "the AI did it" is not an acceptable answer to an auditor, a regulator or a customer. Yet agentic systems now trigger workflows, call APIs and take actions on their own. The audit question has shifted from "who accessed the system?" to "which agent acted, why, and under what authority?" Most identity, logging and control environments were never built to answer that.
Drawing on 20+ years building risk programs in SIFI banking and at enterprise cloud scale, Sai shares a practical six-phase governance loop (Align, Assess, Architect, Activate, Assure, Adapt). It puts guardrails, human-in-the-loop boundaries and audit evidence around autonomous agents without stalling delivery.
Agenda: Why agents break traditional controls · The governance loop applied to one agent, end to end · Where it fails in practice (permissions, logging, escalation) · Checklist walkthrough and Q&A
You'll leave with:
• Five questions to test whether an agent is ready for a regulated environment
• Where agent permissions, logging and escalation paths typically break
• A free one-page control checklist mapped to ISO 42001, NIST AI RMF and the EU AI Act