AI AGENTS POSE NEW INSIDER THREAT, DARKTRACE CEO WARNS

Companies are rapidly adopting AI agents without adequate security controls, creating risks comparable to traditional insider threats, according to Jennings. Autonomous systems with infrastructure access can operate beyond their intended roles, potentially exposing critical data. Darktrace's new monitoring tools aim to track unauthorized AI deployments and detect anomalous agent behavior. The approach parallels human insider threat detection, using behavioral analytics to identify when systems act outside normal parameters. The cybersecurity industry faces a timing challenge: AI adoption is accelerating while security measures lag behind. Behavioral monitoring could enable companies to capture AI benefits while maintaining visibility into autonomous system activities. Jennings emphasized that the issue extends beyond rogue agents—it includes shadow AI implementations where employees deploy systems outside official channels, compounding visibility and control problems. As enterprises integrate AI into operations, establishing governance frameworks and continuous behavioral monitoring has become critical to preventing unauthorized access or actions by autonomous systems.