Security Leaders Converge on AI-as-Identity Framework to Manage Agentic Risk

At an Axios Live Expert Voices roundtable on April 21, 2026, cybersecurity leaders from Mastercard, IBM, Keyfactor, and Okta aligned around treating AI agents as workload identities within existing IAM frameworks rather than creating separate security stacks, with Okta's AI security head calling it a 'moment of rebirth' for the industry.

This practitioner consensus provides a pragmatic path forward for enterprises deploying agents at scale: extend identity threat detection, RBAC, and authorization policies to AI workloads rather than waiting for net-new agentic security categories to mature.

CISOs building AI governance programs should audit whether their IAM platforms can issue, revoke, and monitor credentials for AI agents with the same rigor applied to human and service accounts, and whether they can enforce least-privilege access as agents spin up and down dynamically.