Cisco Expands IAM Portfolio to Secure Agentic AI Traffic
Cisco is expanding its Identity and Access Management (IAM) portfolio to address the growing issue of non-human identities in modern enterprises. These non-humans, including service accounts, API keys, workload tokens, SaaS app grants, and AI agents, are becoming increasingly common as agentic AI rises.
According to Cisco's VP of Product, Identity, Matt Caulfield, non-human identity management is difficult because these entities don't sign in like humans do. They exist in the chaos of machine-to-machine sprawl, outnumbering human identities by 25x to 50x in modern enterprises.
Cisco has made several targeted acquisitions to help understand and manage agentic AI, including Splunk for observability, Galileo Technologies for preventing behavioral drift, and Astrix Security for identifying and understanding agents. These capabilities will be integrated into Cisco Cloud Control, allowing organizations to control how agents access different apps.
Caulfield emphasizes the importance of context in managing non-human identities, as it defines operational purpose, maps infrastructure dependencies, and clarifies risk exposure. He also notes that general-purpose agents are harder to secure due to their breadth of coverage and behavioral span.