Cisco Expands AI Defense and Introduces AI-Aware SASE for Secure Agentic Workflows

Cisco Systems, Inc. has announced a broad expansion of its AI security portfolio, introducing new AI Defense capabilities and AI-aware enhancements to its Secure Access Service Edge (SASE), announced in a press release at Cisco Live EMEA in Amsterdam.

The latest update to Cisco AI Defense adds AI supply chain governance and runtime protections designed to secure agentic tools and interactions. Key features include an AI Bill of Materials for visibility into AI assets, an MCP Catalog for managing model context protocol servers, and advanced algorithmic red teaming for multi-turn testing. Real-time agentic guardrails have also been introduced to detect manipulation and unauthorized tool use. The platform now integrates with NVIDIA NeMo Guardrails, supporting real-time AI protection in production environments.

Cisco’s SASE platform gains new functions for managing and securing AI traffic. These include AI traffic optimization to maintain performance during high-demand periods, visibility and policy controls for MCP communications, and intent-aware inspection to analyze agentic actions. The system unifies policy enforcement across SD-WAN and SSE to simplify governance as AI adoption expands.

Additionally, Cisco unveiled IOS XE 26, the latest version of its network operating system, which introduces full-stack post-quantum cryptography protections. The update supports the company’s 8000 and C9000 Series routers and switches, as well as new 8100 Series models for smaller businesses, providing resilient and encrypted connectivity for AI-driven operations.