Cisco doubles down on Agentic Security

 

By Jim Lundy

Cisco doubles down on Agentic Security

Cisco recently announced a significant expansion of its security portfolio designed to address the unique risks of the autonomous agent era. These updates focus on securing the AI supply chain and governing agentic interactions through AI-aware infrastructure. This blog overviews the Cisco Agentic Security news and offers our analysis.

Why did Cisco announce Agentic AI security capabilities?

The shift from basic chatbots to autonomous agents that execute tasks across hybrid environments has created a massive new attack surface. Traditional security tools struggle to interpret the semantic intent of agent-to-tool communications, leaving a gap for prompt injection and tool manipulation. Cisco is positioning itself to close this gap by integrating security directly into the network fabric and the AI application layer. By launching features like an AI Bill of Materials and Model Context Protocol governance, Cisco aims to provide the visibility required to manage digital labor at scale.

Analysis

This announcement is a major validation of the Aragon Research vision for Agentic Identity and Security Platforms. We have long argued that as agents take on critical enterprise roles, they must be treated as first-class identities with specialized governance frameworks. Cisco’s move to provide runtime guardrails and intent-aware inspection moves the market away from static security toward the dynamic, context-based protection we believe is necessary for autonomous systems.

The impact of this news is twofold. For Cisco, it cements its position as a platform provider that can bridge the divide between networking and AI safety. For the broader market, it sets a high bar for competitors who must now move beyond simple “AI-enabled” security to “agent-aware” security. Cisco is effectively arguing that you cannot secure agentic workflows without deep visibility into the network paths they travel and the protocols they use to communicate. This shift will likely force other security vendors to accelerate their own development of specialized agentic defense layers to avoid being relegated to legacy status.

The Rise of Agentic Scurity and Identity Platforms

Cisco follows others, such as industry Pioneer Britive in making this move. In fact, Aragon feels that a majority of Security Platform as well as Identity providers will need to adapt to AISP, given the explosive rise of AI Assistants and AI Agents.

For enterprises, this goes beyond an application trust layer. It is about making sure that internal and external agents are identified and verified before granting access.

What should enterprises do about this news?

Enterprises should evaluate these new offerings as they move from pilot programs to production-grade agentic systems. It is no longer enough to secure the human user; the focus must expand to the automated agents that act on behalf of the business. Organizations should consider how Cisco’s AI-aware SASE and post-quantum cryptography fit into their long-term data protection strategy.

We recommend that security architects conduct a thorough audit of their current identity stacks to ensure they can support the high-velocity, high-volume nature of agentic interactions. Additionally, enterprises should Agentic security approaches of Cisco  vs others including Britive, Google and Microsoft.  While the market is still early, it is not too early to adopt a platform and get started. Work with providers and get a pilot project into sandbox production. Test to verify that unauthorized agents do not gain access.

Bottom Line

Cisco is taking a necessary step toward securing the next generation of digital labor by embedding agentic protections into its core architecture. This move aligns perfectly with the need for robust identity governance in an world of autonomous AI. Enterprises must move quickly to adopt these agent-centric security models or risk facing significant governance and exposure gaps as their AI adoption scales.