Rogue AI: Leaders Confident in Detection, Not Containment

Many IT and security leaders are confident they can detect when an AI agent malfunctions, but few can quickly trace and contain its impact. Nine out of ten leaders believe they can find malfunctioning agents. However, only 26% say they can trace the downstream impact within minutes. Over 45% state it would take hours to understand the full impact of an agent incident. This delay between detection and mitigation is a significant problem. Jeffrey Collins, CEO of WanAware, notes that malfunctioning agents can cause major outages and data breaches very quickly. Kevin Paige, a field CISO, emphasizes that agents operate at machine speed. The damage can spread across systems before anyone can pinpoint the agent. Often, organizations learn about rogue agents from customers or broken systems, not their own tools. This matters because the inability to quickly contain a malfunction can damage trust and slow down AI adoption.