Seeing the Agents in the Room: Five Steps to Governing Claude Code and Claude in the Enterprise
Most organizations are using AI. Few can tell you who is using it, what it's doing, or what credentials it's acting on. Agentic tools like Claude Code and Claude for Work read files, execute commands, invoke external services, and operate autonomously — introducing risks that chatbot-era governance was never built to detect. The incidents are already here, from production databases deleted by coding assistants to zero-click prompt injection exfiltrating data through enterprise copilots.
In this session, ReliaQuest’s Chief Scientist, Brian P. Murphy walks through five practical steps to gaining visibility over agentic AI in your enterprise — covering how to enforce policy, collect the right telemetry, deploy detections in your first week, and prepare for a regulatory landscape that's converging fast on AI accountability.
What you'll learn:
What makes agentic AI fundamentally different from chatbot-era tools — and what that means for your governance model
Five steps to AI visibility: from policy enforcement and permission controls to week-one detections that surface shadow AI, credential risk, and runaway autonomy
Where regulators are heading on AI risk management and logging — and how to build a governance posture that's ready now
