Rethinking Identity Security for the Age of AI
As artificial intelligence agents take on more autonomous roles across enterprise environments, organizations must rethink identity security around accountability, attribution and real-time governance, according to Vibhuti Sinha, chief product officer at Saviynt.
Sinha explained how the rapid growth of non-human and agentic identities is exposing the limitations of traditional identity and access management models. He said enterprises need continuous visibility, runtime authorization controls and stronger ownership frameworks to secure AI agents as they learn, adapt and act independently.
In this video interview with ISMG, Sinha also discussed:
- Why auditability, attribution and accountability are becoming foundational principles for securing AI agents
- The need for continuous governance, ownership management and real-time visibility into non-human identities
- How runtime authorization and access gateways can enforce least-privilege controls for autonomous agents