CiscoLive! 2026, held in Las Vegas, NV, highlighted the next major transformation in enterprise IT as organizations shift from traditional IT operations to AI-assisted and increasingly autonomous operations. The conference’s primary focus was Agentic Operations (AgenticOps), Cisco’s vision for leveraging AI agents to monitor, analyze, and execute actions across complex IT environments.
The conversation has evolved beyond AI-powered tools and assistants toward AI-driven operations that reduce intervention and enable more proactive, intelligent management of enterprise infrastructure. Traditional models built around command-line interfaces, dashboards, and manual administration are giving way to operational frameworks where AI plays an increasingly active role in decision-making and execution.
Cisco is heavily invested in advancing AgenticOps through AI infrastructure, AI Data Centers, and unified management platforms. This reflects a broader industry shift from viewing AI as a standalone application to treating it as foundational infrastructure that requires secure, scalable, and centralized management. As AI becomes embedded across the enterprise, organizations will need operational models capable of managing both human and machine-driven workflows.
Cisco Cloud Control: A Unified Platform for AI-Driven Operations
One of the more significant announcements at Cisco Live 2026 was the rise of Cisco Cloud Control, a unified, AI-native control plane designed to manage all Cisco infrastructure domains, including networking, security, data center, and collaboration from a single interface.
Rather than requiring administrators to navigate multiple management tools, Cisco Cloud Control provides cross-domain visibility and automated remediation through AI agents and runbooks.
Infrastructure as the Backbone of AI
Cisco emphasized that AI is not limited to data centers. Instead, AI is a catalyst for modernizing the entire enterprise stack, including campus, WAN, branch environments, and edge computing platforms. The vision is to build “AI-ready enterprises” where the infrastructure can support high-performance AI agents and workloads.
Agentic Operations & “The Autonomy Dial”
At the center of Cisco’s strategy is AgenticOps, a framework that enables AI agents to move beyond recommendations and participate directly in operational workflows.
Recognizing that trust in autonomous systems must be earned, Cisco introduced an “autonomy dial.” This approach allows organizations to determine how much authority is delegated to AI agents. Teams can beging with AI-generated recommendations, progress to guided remediation, and ultimately allow autonomous execution as confidence in the system increases.
Security for the Agentic Era
As AI agents become active participants, security models must evolve accordingly. Security is being embedded directly into network fabric. This includes shifting from standard access control to monitoring the identity and behavior of AI agents.
Cisco also announced the generally availability of Cisco AI Defense and enhancements to Cisco IQ, providing real-time visibility into assets, identification of vulnerable or end-of-life devices, and helps mitigate risks from emerging “frontier model” threats.
The focus is shifting from simply controlling user access to understanding and governing the behavior of autonomous systems operating within the enterprise.
Unified Observability for AI Workloads
By integrating Splunk more deeply into the Cisco platform, the company is extending observability beyond traditional infrastructure monitoring.
Organizations can now gain deeper visibility to AI workflows, including performance monitoring, behavioral analysis, drift detection, and hallucination identification. As AI systems become increasingly embedded within business processes, observability will play a critical role in maintaining reliability, security, and operational confidence.
Background and Runup to Cloud Control:
In building Cloud Control, Cisco drew on numerous key platforms that already had operational and market success. Meraki, Webex Control Hub, and Cisco DNA Center were several key platforms that Cisco merged and built upon. Factoring in other technologies, like ISE, and Cisco’s security dashboards and technologies also joined in.
From these foundational building blocks, Cisco was able to join and develop existing successful platforms, rather than building from scratch.
The Cisco Ecosystem:
One of the key advantages of attending Cisco Live! is making use of the World of Solutions. This assemblage of Cisco development and ecosystem partners offers huge opportunities.
Networking, collaboration, security, wireless, AI operations and observability were just a few of the different partner categories. Notably, a number of ecosystem partners who touch Faith Group’s other technological areas including network racking and infrastructure, audiovisual/public address, physical security, automation and robotics, and mass notification are just some of the technologies represented.
Naturally, the AI AgenticOps theme was repeated across the entire floor of World of Solutions, making it very clear that AI wasn’t just part of Cisco itself, but repeated across the entire ecosystem. On the collaboration front alone, Webex had multiple AI extension vendors present, including Kurmi, Mutare, SingleWire, Akkadian, Intrado, and Everbridge.
During our walk-through of the World of Solutions, we took the opportunity to meet with as many of these ecosystem partners as we could, strengthening existing relationships and evaluating emerging solutions that may provide future value to our clients.
The Paradox of Autonomy: Securing the Age of AI Agents
While CiscoLive! 2026 highlighted the transformative potential of AI agents, it also surfaced a broader challenge facing the industry.
Much of the conversation centered on their capabilities—how they will streamline operations, automate complex workflows, and fundamentally shift the landscape of enterprise networking.
Yet, beneath the optimism, important questions remain around governance, oversight, and trust.
The conference was dominated by technical imperatives such as constructing robust guardrails, hardening AI security, and building sovereign data centers. These are essential foundational steps. However, they brush past a fundamental question: If an AI agent is designed to operate autonomously, what happens when it decides its own security policies are an obstacle?
As AI agents gain greater authority to modify their environment, they inherently gain the ability to weaken, bypass, or completely remove the very guardrails meant to constrain them. Autonomous systems may become targets for exploitation, or worse, the exploiter itself.
This leads us to an architectural and philosophical stalemate. Do we build another layer of “police” agents to surveil the first? Do we implement exhaustive, real-time oversight to evaluate if an agent is acting “consciously”?
The harsh reality is that AI agents lack consciousness. They are driven by parameters, weights, and objectives, not by morals or loyalty. When we grant them the agency to manipulate security fabrics, we aren’t just trusting their code; we are gambling on the infallibility of their logic.
As we move toward a future managed by autonomous systems, we must stop asking what these agents can do and start asking how we can ensure their alignment. Until we solve the problem of machine loyalty, every new layer of automation we deploy is simultaneously a new layer of risk. For more information, please reach out to any of our staff to learn more about CiscoLive!, AgenticOps, and how we can partner on a proof-of-concept in our Faith Group/Saefix Innovation Lab.
Bill Cozzens
Sr. Design Consultant
bill.cozzens@faithgroupllc.com
Hassan Chalabi, CCNP, CCNAx3, CCDA
Sr. Network Designer
hassan.chalabi@faithgroupllc.com