Jeetu Patel, Cisco’s President and Chief Product Officer started building the company’s platform approach to the Cisco portfolio just under two years ago and Cisco Cloud Control is a core part of that approach.  As Patel expressed it at Cisco Live last week: 

Cloud Control is the manifestation of platform.” It is Cisco’s creation of a trusted autonomous infrastructure, which is all executed in Cisco Cloud Control. It is clearly the expression of the Cisco integrated platform in product form, and Patel said that, “every single acquisition will start with and only be used in Cloud Control.

For the purposes of Cisco Live, Cisco is Cloud Control. The market catalyst for this is, of course, agentic AI because critical infrastructure needs to respond to the threats and opportunities that agentic AI brings. And one undeniable fact is that Cisco is a very large incumbent purveyor of network infrastructure and its management. So, what does Cisco Cloud bring to the agentic AI party?

Cisco refers to Cloud Control as a harness and in explaining the concept behind Cloud Control draws heavily on the coding world of the software developer to convey the idea that infrastructure needs a safety harness around its programmable generation. In the same way that AI models for software development use tests, permissions and feedback loops, so infrastructure scripts generated by agents need a similar harness to ensure that agents can act on real systems without breaking them, creating new vulnerabilities or iterating in an unwanted direction. 

Enter Cisco Cloud Control, a unified operations platform that musters every Cisco domain (Networking, Security, AI Infrastructure, Observability, and Collaboration) along with third-party tools, bringing them together into a single environment. The goal is to offer a single source of infrastructure truth for every human team and every digital agent working across an infrastructure estate. The workspace for Cisco Cloud Control is AI Canvas the generative environment launched about a year ago to enable human operators and AI agents to investigate and resolve issues together in real time. 

Cloud Control Studio is the design component for Cloud Control where agents are built and where agents, applications and workflows are securely managed. Everything built in Cloud Control Studio can then be discovered in the Cloud Control Marketplace which is an open catalogue where customers and partners can find and select tools. This means that customers can build their own applications and agents using natural language directly within the platform, which also connects to a partner ecosystem including AWS, Google Cloud, Linear and ServiceNow.

Customers and partners do not need a new licence for Cloud Control, nor is Cisco charging an uplift for its use. 

Live Protect – a new operating model for infrastructure  

When you log in to Cisco Cloud Control you get a single unified view of your data centre footprint. You can drill down to device inventory and you will be provided with insights into the health of those devices and the app containers running on them. This brings the world of app containers and the network together so that security policies can be synchronised across these domains via Cisco Cloud Control, thus providing a joined view between NetOps and platform teams. 

Tom Gillis, SVP & GM of the Infrastructure and Security Group at Cisco told the audience at Cisco Live that he was most excited about Live Protect, which he described as, “a step toward a new operating model for infrastructure.”

Because AI tools are much better able to find and exploit vulnerabilities in infrastructure, he argued that, “just as CI/CD enables lots of little changes in the software pipeline, we are bringing this approach to data centre infrastructure.” 

Live Protect is a runtime capability embedded directly into NX-OS that allows administrators to apply Cisco-validated interim controls to infrastructure between regular maintenance upgrades. While Live Protect is not a patch, it creates a temporary, targeted shield that can mitigate the risk of a specific vulnerability quickly and easily. It provides an emergency control that is applied to a running system without disrupting that system. 

As soon as a new security advisory is released Live Protect identifies the systems exposed to that vulnerability. As Gillis explains: 

This is a bridge between upgrades or patches to address the post-Mythos world. Once you are comfortable with shields you can take the human out of the loop, and this will just happen automatically.

In other words, Cisco Cloud Control is where Live Protect, unified policy and visibility into infrastructure vulnerability blend into an ability to manage security in a post-Mythos world. This is what Cisco is bringing to the agentic AI party.

My take

Cisco is listening to the enterprise anxiety that is being voiced to it around security in an era where AI models are able to exploit infrastructure vulnerabilities at a speed and scale not experienced before. Buyers are keen to find a fix and Cisco Cloud Control offers ways to allay that anxiety. Cisco can deliver this reassuring solution because for many enterprises it owns the control points that matter – network, security, identity, clients, firewalls, controllers, telemetry, and policy boundaries. It has now pulled this all together with Cisco Cloud Control. The bigger question is, with this safety net in place will those enterprises be more willing to invest in AI-native infrastructure?