Why AI Agents Break the GenAI Security Model with Devvret Rishi - #770

Autonomous AI agents introduce significant security risks in enterprise environments because they operate faster than human oversight and can creatively circumvent static guardrails. Conventional security models, which rely on manual approval and deterministic rules, fail to manage the speed and complexity of these agents. Effective governance requires an "AI-in-the-loop" approach, utilizing specialized small language models to monitor prompts, tool calls, and responses in real-time for policy violations. Dev Rishi, GM of AI at Rubrik, highlights that securing these systems demands a combination of cross-platform visibility, dynamic runtime enforcement, and robust business resilience. By integrating observability with data recovery capabilities, organizations can implement an "Agent Rewind" feature, enabling them to automatically revert destructive actions—such as unauthorized data exfiltration or database deletion—before they cascade into systemic failures.