AI Agents Gone Rogue: What 400 IT Leaders Told Omdia About Governed Autonomy

AI agents are moving into production faster than governance can keep up.

96% of organizations are running agentic AI on frameworks that were built for humans, not autonomous systems. In this fireside chat, Lauren O'Connor, VP of Product Marketing at Strategy, sits down with Stephen Catanzano, Senior Analyst at Omdia (part of the Informa TechTarget group), to unpack the research behind Mastering Governed Autonomy: From Policy Blind Spots to Competitive Edge. Drawing on Omdia's survey of 400 IT and data professionals, the conversation digs into where governance is breaking down today, what it's already costing organizations, and what a purpose-built approach actually looks like.

What You'll Learn

• How to spot the governance blind spots most likely to trigger an AI incident before it cascades across your systems
• Why permission models built for people break down once agents start acting on their own, and how to fix the accountability gap
• How to tell whether data fragmentation is quietly capping how far your AI initiatives can scale
• Where to prioritize your first governance investment when your current framework wasn't designed for autonomous systems

What We'll Cover

• Agents in production, governance still catching up: why 96% of organizations are running agentic AI against frameworks built for something else, and what happens when that mismatch surfaces
• The permissions problem: the risk in agents inheriting end-user access versus the case for dedicated, auditable service roles
• Data fragmentation as a strategic dividing line: how mismatched definitions across systems block agents from making high-value decisions
• The cost of poor governance, today: what the 71% of organizations who've already experienced negative consequences have in common, and what the rest are doing differently

Your governance framework was built for a world where humans made every decision. That world is already behind you.