Cisco
Why They Matter
Cisco is setting the enterprise security standard for agentic AI. Its Duo Agentic Identity product ā announced at RSAC 2026 ā extends zero-trust identity controls to AI agents, treating them as first-class security principals rather than trusted extensions of human accounts. Cisco's enterprise reach means its frameworks become procurement requirements for large organizations. Its RSAC 2026 warning that "one wrong agentic AI action could cause irreversible damage" has become the defining public statement on agent risk from a major infrastructure vendor.
Background
Cisco built the core infrastructure layer of the internet ā routers, switches, firewalls, and networking hardware that power enterprise connectivity. It expanded into security through acquisitions and organic development. The shift to agentic AI has exposed a new attack surface that Cisco's traditional security tools weren't designed for: AI agents that act with human-level authority, execute code, move data, and interact with external systems autonomously. Cisco's RSAC 2026 announcements represent the first major enterprise networking company to build a dedicated security framework for the agent layer.