AI Agents: New Identity Security Challenge for Businesses

AI agents are creating a new identity security problem for businesses. As companies deploy these agents, they're expanding beyond human identities into machine and agentic identities. This creates security, governance, and compliance challenges. Rishi Kaushal, CIO of Entrust, says AI agents need to be treated as first-class identities, with accountability, auditing, and oversight. They access systems and make decisions, so organizations need visibility into their actions. Kaushal states that every agent should have its own identity, defined permissions, and clear accountability. The issue is growing as companies deploy AI assistants and autonomous systems without governance frameworks. This leads to agent sprawl and increased risk. Zero-trust principles, like continuous verification and least-privilege access, must extend to AI agents. Without these controls, organizations face risks from poorly managed agent permissions. This matters because it highlights a critical new area for businesses to secure their digital operations.