Cisco launches Cloud Control platform where AI agents and humans collaborate on IT operations

Cisco AI Canvas is here: the workspace for agentic operations

A year ago, at Cisco Live, we previewed Cisco AI Canvas, an agentic workspace where IT teams and AI agents investigate and resolve issues across every domain. It was built for teams that need to move from alert to evidence, from evidence to decision, and from decision to resolution faster. During beta testing, AI Canvas was available in Meraki and Splunk. Today, Cisco AI Canvas is moving into Controlled Availability as an integrated part of Cisco Cloud Control rather than in the individual products themselves. This release reflects months of beta feedback from operators, architects, and IT leaders across industries who have used AI Canvas in real environments. What AI Canvas was built to solve Cisco has spent four decades building category-defining products across networking, security, compute, observability, and collaboration. Each product is powerful on its own. The next opportunity is what becomes possible when they work together as one operational platform. A single operational question often spans many domains. Why is an application slow? Why is call quality poor? Did a recent firewall change affect user experience? Operators often become the integration layer, moving across tools and assembling the incident story by hand. The signal is usually there. The challenge is correlation, not data, which is why mean-time-to-resolution (MTTR) gets stretched to hours and sometimes days. Cisco Cloud Control provides the foundation that brings Cisco products into one operational environment, giving teams a shared way to view their environment, get correlated alerts across products, and take action with AI assistance. AI Canvas is built on that foundation, as a workspace where the shared context becomes agentic action. Inside an AI Canvas investigation AI Canvas changes the operating model from manual investigation to agent-led investigation, with the operator in control. Operators ask questions in natural language. From a single prompt, AI Canvas runs a structured multi-agent investigation. It reads the question, identifies the domains involved, builds a plan, and dispatches specialized agents to gather evidence in parallel. AI Canvas synthesizes the findings into one sourced answer, with the reasoning trail visible so operators can defend the conclusion or hand it off cleanly. Default Mode powers the day-to-day experience. When operators need to check the health of a site, look at recent activity, or pull a quick view of an asset, Default Mode returns answers quickly with the relevant context. Most operational questions live here. For more complex incidents, this Controlled Availability release brings something new. What is new in Controlled Availability This release introduces capabilities that make AI Canvas more useful for daily operations and more trusted for complex investigations. 1. Deep Reasoning Mode is built for harder, multi-domain problems where teams need a clear, defensible answer. AI Canvas creates a full troubleshooting plan up front, grounded in best practice. The operator reviews, revises, or approves the plan before any agent runs. As the investigation proceeds, every step, finding, and conclusion is sourced and supported by evidence. Operators end up with an investigation that is easy to follow, easy to defend, and easy to hand off, especially when getting the answer right matters most. Default Mode keeps everyday operations moving. Deep Reasoning Mode steps in when issues require deeper analysis. 2. Interactive generated widgets give operators the exact view they need for the question they asked, including topology maps, performance charts, summaries, and reports drawn from live data. Operators can click in, drill down, export, or push a widget into the next step of the investigation. 3. Multimodal context lets operators bring in the visual evidence they already rely on. Screenshots, dashboards, RF heatmaps, topology diagrams, and error images can become part of the investigation alongside live telemetry. The result is more complete investigations because AI Canvas reasons over the same evidence the operator is looking at. 4. Knowledge base built in lets teams bring runbooks, SOPs, and policies directly into AI Canvas. AI Canvas starts with Cisco best practice and becomes more specific to each customer's environment as approved knowledge is added. The result is AI that follows the team's standards, with investigations aligned to how the organization actually operates. 5. Board Library gives operators Cisco-curated starting boards for common scenarios, including security posture reviews, wireless troubleshooting, compliance checks, and application performance investigations. Each board is structured so teams can run a meaningful investigation immediately rather than start from a blank page. Each of these capabilities exists because operators told us what an agentic workspace really needed. Shaped by operators across industries This release has been shaped by alpha and beta customers across industries - financial services, healthcare, retail, manufacturing, technology and more. Their teams ran AI Canvas against real environments, real incidents, and real operational patterns. They helped us refine what operators need from an agentic workspace: clear plans, sourced findings, useful visualizations, persistent context, faster investigation, and human control. That feedback is visible throughout this release. We are grateful to every customer and operator who helped shape this milestone. Designed for how teams work AI Canvas is designed for how operators actually work. It is where humans and AI agents come together in one shared workspace, working from the same evidence, the same context, and the same proposed actions. Multiple teammates can join the same investigation in real time, while AI agents handle the heavy lifting of gathering context, correlating signals, and proposing next steps across domains. Findings, decisions, uploaded files, and approved actions stay with the investigation across handoffs, shift changes, and escalations, so the next operator picks up where the last one left off rather than starting cold. That continuity is what helps reduce MTTR. Instead of repeating work or rebuilding context, teams keep the work moving across people, shifts, and the AI agents working alongside them. Customize & extend with Cloud Control Studio Agent Builder in Cisco Cloud Control Studio was also announced at this year's Cisco Live. It is a new part of Cisco Cloud Control, designed for customers and partners who want to extend the platform with their own tools, agents, and operational knowledge. In Agent Builder, customers can do three things: * Connect third-party tools. Through native integrations or open Model Context Protocol connectivity, tools beyond Cisco can become part of how AI Canvas reasons across the environment. * Build custom agents. Customers can build agents tailored to their own operational needs, such as configuration drift monitoring, compliance checks, escalation workflows, or recurring incident investigations. Agents join investigations live inside AI Canvas. * Turn knowledge into reusable skills. Runbooks, SOPs, and procedures become reusable agent skills that any authorized agent can call during an investigation. Teams can also discover, share, and adopt agents and integrations through the Cloud Control Marketplace, the third-party ecosystem inside Cloud Control. Marketplace gives teams a curated way to extend the platform with capabilities built by Cisco and partners. Anything connected, built, or adopted through Cloud Control Studio and the Marketplace is usable in AI Canvas. The more tools, agents, and skills added, the more capable the workspace becomes for the customer's specific environment. Come see it at Cisco Live Cisco AI Canvas, alongside Cisco Cloud Control, is entering Controlled Availability for United States commercial customers. If you are at Cisco Live, come see Cisco Cloud Control with AI Canvas at the booth for a live demo, or in sessions throughout the conference. Session catalog for AI Canvas here. If you are not attending, visit the Cisco AI Canvas webpage to stay up to date on the latest product innovations. Some products or features described may be in various stages of development and offered on a when-and-if available basis.

Cisco Cloud Control: The Secure Harness for the Agentic Era

Every few decades, infrastructure gets a new abstraction. We went from racking hardware to virtualizing it, then from clicking through consoles to writing infrastructure as code. Each shift changed who could build, how fast they could move, and what was possible to defend. The next abstraction isn't coming. It's already here. As Jeetu Patel captured, agentic AI is kicking off a networking supercycle -- the era of managing critical infrastructure at human scale is over. What comes next has to be built for a workforce that is no longer entirely human. Agents are about to become a permanent part of how enterprises run. Real, capable, reasoning agents that observe infrastructure, modify it, defend it, and extend it. The real question: is your infrastructure ready to let them in safely? That's the question Cisco Cloud Control was built to answer. Above: Cisco Cloud Control homepage From infrastructure as code to infrastructure as a harness Think about what made Codex and Claude Code feel like a step change for software developers. It wasn't just the model, but the harness around the model: the repo, the terminal, the tests, the permissions, the feedback loop. The model was powerful, but the harness is what made it useful. Infrastructure needs the same thing. Terraform, Ansible, and Python scripts made infrastructure programmable, but they all assumed a human was writing the logic, stitching the systems together, and deciding what to change. That assumption no longer holds. Agent-native infrastructure demands a different abstraction. It needs its own harness: a safe, governed, observable control surface that lets agents act on real systems without breaking them, exposing them, or taking them somewhere they shouldn't go. That is what Cisco Cloud Control was made to do: the unified operations platform that brings every Cisco domain (Networking, Security, AI Infrastructure, Observability, and Collaboration) and your third-party tools into one environment, with one login, one view, and one operating model for every team and every agent working across your estate. This isn't a single pane of glass; glass is passive. Cisco Cloud Control is active execution, with policy and identity built into the control path itself. What the harness actually does A harness is a concrete set of capabilities that must be real, or none of this works. Cisco Cloud Control gives agents -- and the humans who direct them -- six things at once: * Trusted access to routers, switches, controllers, firewalls, clients, users, workloads, and applications that make up the modern enterprise. * Normalized APIs and MCPs for observing and managing every domain. * Identity, policy, and zero trust built directly into the control path, rather than bolted on top. * Telemetry and operational context that show what is actually happening on the ground, in real time. * Enforcement points that can block, isolate, reconfigure, or remediate at runtime. * Governance that makes every agentic action transparent, auditable, bounded, reversible, and subject to human approval. This is what "real" looks like. Six capabilities, working as one -- the harness that makes agents safe to deploy, and powerful enough to matter. Where the work happens: AI Canvas The harness is the engine. The workspace is where you actually drive. Cisco Cloud Control's workspace is AI Canvas -- the multiplayer, generative environment where human operators and AI agents investigate, correlate, and resolve complex issues together in real time. Operators see the same live evidence that agents do. Agents build investigation plans, pull data across domains, and drive resolution end to end. Context persists across shifts and escalations, so nothing is lost and nothing is repeated. What used to require a war room now happens in a workspace. Cisco Cloud Control is the platform; AI Canvas is where the work happens -- learn how it has evolved since we first unveiled it last year. Above: AI Canvas workspace Where customers build: The Studio in Cloud Control AI Canvas makes agents useful. Cloud Control Studio is where you build them. Cloud Control Studio is the design space -- the factory floor -- where customers and partners build and secure the agents, applications, and workflows their business depends on, on top of a substrate that's already wired into Cisco's data, policy, and control plane. Inside Studio, two capabilities do the heavy lifting: * Agent Builder -- where AI agents for Cloud Control are created. Customers and partners can build, train, bring, and secure their own agents. * App Builder -- where custom apps and workflows are developed for Cloud Control. Users build and publish from natural-language prompts, with built-in agentic coding assistants including OpenAI Codex. Everything built in Cloud Control Studio can be discovered in the Cloud Control Marketplace: the open catalog where customers and partners find and extend what's possible, so teams can find the right tools for the right task. Customers can build their own applications and agents using natural language directly within the platform, which also connects to a large ecosystem including AWS, Google Cloud, Linear and ServiceNow. Above: Cisco Cloud Control Marketplace The customer unlock Picture this. An IT admin walks in on Monday morning to discover that the operations team has begun rolling out a new class of AI-enabled inspection cameras across the manufacturing floor. The devices are already on the network, but they sit outside every existing management tool. The vendor has an API, but no integration exists yet. In the old world, the safe answer was no - wait for the roadmap, wait for the connector, wait for someone else to build it. In the new world, the answer changes. The admin opens a coding harness, builds a small Cisco Cloud Control application, connects to the device API, maps each camera to network location and identity context, pulls telemetry, and exposes policy actions through Cisco's governed control surface. In hours, not quarters, those devices become observable, manageable, and defensible -- under the same policy model as everything else in the estate. That same pattern works for a new SaaS app, a custom internal system, a factory controller, a medical device, or a partner-managed appliance. Anything the business throws at IT before the vendor catalog catches up. The real unlock: customers can build their own workflows and inherit Cisco's APIs, policy model, observability, enforcement, and governance for free. Why this matters now Powerful models are making software creation faster than it has ever been, and that is a gift for productivity. It also reshapes risk. If models can generate code at this speed, they can generate vulnerable code at this speed - new applications, new attack surfaces, and new runtime behaviors that traditional security workflows weren't designed to catch in time. The window between vulnerability discovery and exploitation has collapsed from weeks to minutes. What customers need is the ability to build custom shields around their own applications and infrastructure: agentic defenses that observe behavior, understand context, detect exploit patterns, reconfigure controls dynamically, block attacks at runtime, and verify the mitigation actually worked. This is where Cisco Cloud Control becomes more than an operations platform -- it's the command center for a post-Mythos security posture, the place where Live Protect, unified policy, and full vulnerability visibility come together. Cisco can deliver this because Cisco owns the control points that matter - network, security, identity, clients, firewalls, controllers, telemetry, and policy boundaries - and the harness sits on top of all of them. The superhuman admin Let's be clear: none of this is about replacing the people who run infrastructure. It's about making them dramatically better at it. That's what AgenticOps means in practice -- Cisco's operational system for the AI era, where governed AI agents monitor, reason, and act across your infrastructure with humans firmly in the loop. AgenticOps is the capability. Autonomous infrastructure is the outcome. Cisco Cloud Control is where it all comes to life. In practice, that means coding tools help the admin write the custom app, Cisco Cloud Control gives that app secure access to the infrastructure, Cisco APIs and MCPs expose the operational capabilities, Cisco policy and identity govern what the app can do, and Cisco enforcement points make the app useful in the real world. The IT admin becomes a builder. The security admin becomes a rock star. The platform team starts shipping at the speed of the business. That's what it looks like when strategy turns into reality. The shift is happening now Cisco has spent four decades building best-in-class solutions across every domain that matters in enterprise IT. Meraki. ThousandEyes. Security Cloud Control. Nexus. Intersight. Webex. Each one a leader. Our depth is unmatched -- and there's more to come. We innovate from the silicon to the application -- making our own chips, building the systems they power, writing the software that runs on them, and providing the security and observability that watches all of it. The opportunity lives in what becomes possible when every domain works together -- when shared inventory, unified topology, and a single environment turn the portfolio into a platform. That's the architectural choice underneath Cisco Cloud Control: tightly integrated, loosely coupled, and built for an open ecosystem. The more Cisco you run, the more powerful the platform becomes -- and the heterogeneous environment you already have keeps working alongside it. We're building the secure harness that lets every customer connect agents to their infrastructure, build their own applications on top of it, and defend their environments at the speed AI is moving. That's the shift. Dashboards giving way to agentic workflows. Infrastructure as code evolving into infrastructure as a harness. Fixed product features expanding into customer-built infrastructure apps. Static controls becoming custom runtime shields. Cisco evolving from a portfolio of products into the secure harness for enterprise infrastructure in the agentic era. This is Cisco's next big thing. We're glad you're here for it.

Cisco's new cloud platform aimed at securing AI infrastructure

Cisco's new cloud platform aimed at securing AI infrastructure Cisco Systems Inc. today introduced a sweeping set of products and services designed to help enterprises manage, secure and automate increasingly complex information technology environments as artificial intelligence agents become embedded in corporate infrastructure. The announcements, made at Cisco Live in Las Vegas, are "the most consequential that Cisco has made in many years," said Jeff Schultz, senior vice president of portfolio strategy for Cisco's product organization. The anchor is Cisco Cloud Control (pictured), a new platform that unifies networking, security, observability, infrastructure and collaboration management into a single operational environment for both human administrators and AI agents. Cisco said the platform addresses an evolving scenario in which AI agents function as digital coworkers, creating operational demands that conventional IT management tools were never designed to handle. "We are moving from the age of chatbots to the age of agentic AI, where chatbots are productivity tools that help us do our work and agents are digital coworkers," Schultz said. Unlike chatbots, agents operate continuously, interact with other agents and directly access enterprise systems. Executives said that creates new challenges around networking, security and operational visibility. "They're being deployed everywhere now," Schultz said. "Every action that an agent takes is a combination of a routing challenge, a trust decision and a telemetry event." Cisco executives said enterprises face three major barriers to scaling agentic AI: infrastructure limitations, a growing trust deficit around autonomous systems and an explosion of telemetry data generated by AI-driven operations. Anthropic PBC's Claude Mythos has raised the specter of attacks being coordinated at machine speed, requiring defenses that leverage autonomous agents. Automated operations Cloud Control is the foundation of Cisco's AgenticOps, an operating model that shifts IT and security operations from manual, human-led processes to autonomous, agent-driven orchestration with human oversight. Executives described the platform as a response to fundamental changes in how enterprise infrastructure is managed. "It's not just about humans clicking through dashboards trying to keep up, but a true collaborative operating model where agents are doing the heavy lifting and humans are staying in control of what matters," said DJ Sampath, senior vice president and general manager of Cisco's AI software and platform group. Cisco describes Cloud Control as a single management plane that brings together its networking, security, computing, observability and collaboration platforms. Single sign-on access is provided to a unified data layer called Cisco Data Fabric, based on the Splunk log data analysis platform Cisco acquired two years ago. Executives said the platform combines cross-domain telemetry with domain-specific AI models trained on Cisco's operational data. These include specialized networking and cybersecurity models as well as frontier AI models when appropriate. AI agents automate operational workflows. Cisco said they can identify problems, perform root-cause analysis, recommend and implement fixes, test proposed changes on digital twins of the customer's environment and validate outcomes with human oversight and policy controls. "These agents are explainable, policy aware, permissioned, auditable and constrained by the guardrails that we put in," Sampath said. Cisco is also introducing two major platform components designed to help enterprises build their own AI-powered workflows. Cloud Control Studio provides tools for creating custom agents and applications. Its Agent Builder allows organizations to create agents connected to more than 40 third-party platforms, including the major hyperscale clouds, ServiceNow Inc., PagerDuty Inc. and Google LLC's Wiz through native integrations and the Model Context Protocol. Cisco said App Builder enables users to create applications and workflows using natural-language prompts and integrates with OpenAI LLC's Codex software engineering agent and coding assistant. Cisco AI Canvas is a collaborative workspace where operators and AI agents can investigate and resolve issues using shared operational context and telemetry. Cisco said the environment preserves context across shifts and escalations so operational knowledge is not lost. The company also emphasized openness as a key design principle. "Enterprises are fundamentally dynamic," Sampath said. "Every one of them has vendors other than just Cisco." High-speed security The second major pillar of today's announcement is cloud security. Executives said advances in frontier AI models have dramatically compressed the time between vulnerability discovery and exploitation. "The frontier models have fundamentally changed the operating assumptions that have been in place in the data center for more than a decade," said Tom Gillis, senior vice president and general manager of infrastructure and security. Cisco's response is Live Protect, which applies security controls directly to running infrastructure without requiring software upgrades, maintenance windows or system reboots. The technology is now available on Nexus 9000 switches and is expected to expand across additional Cisco products. "Live Protect allows us to identify vulnerabilities in that infrastructure and apply a compensating control with pinpoint accuracy on a running switch without rebooting that switch," Gillis said. The company also unveiled new capabilities aimed at securing AI agents. DefenseClaw is a security and governance framework for local AI agents such as OpenAI Codex, Claude Code and OpenClaw. The technology scans for vulnerabilities, enforces access controls and integrates with Cisco AI Defense and Splunk for monitoring and governance. Gillis said enterprises must simultaneously protect AI agents from malicious environments and protect enterprise assets from potentially risky agent behavior. "We need to be able to protect these agents from hostile environments in the outside world," he said. "At the same time, we need to protect our assets that are living in the outside world from agents." Cisco is also extending its Zero Trust architecture to AI agents by enabling more granular controls over what actions agents can perform and what resources they can access. The company is further expanding its security portfolio with Agentic SOC capabilities that use AI agents to automate workflows in the security operations center. Cisco said the technology can reduce incident response times from hours or days to minutes. Splunk, which Cisco acquired last year, plays a central role in its new security strategy. "Data quality is a primary constraint for agentic systems," said Kamal Hathi, senior vice president and general manager of Splunk. New Splunk capabilities included federated search, a machine data lake and an AI toolkit designed to help organizations process and analyze telemetry at scale. Hathi said new observability capabilities focus on AI environments, including monitoring of agent behavior, performance, safety and costs. "The cost of AI systems and the use of tokens is fast becoming a top concern for organizations," Hathi said. Quantum equation Cisco is also expanding its quantum-security offerings, saying it is committed to enabling quantum-safe communications capabilities across much of its core portfolio by the end of 2026. New Quantum Ready Assessments help customers identify infrastructure vulnerable to future "harvest now, decrypt later" attacks, in which assailants harvest encrypted information intending to decrypt it later with speedy quantum code-breaking tools. On the services side, the company added Resilient Infrastructure Services to its Cisco IQ support and services platform. The offering combines exposure assessments, infrastructure modernization guidance and defense resiliency planning. "Mythos changes the rule permanently," said Bhaskar Jayakrishnan, senior vice president of engineering for customer experience. "Things are coming at us at machine speed, but the principles don't change. We need to help our customers with the structural resilience so that they can withstand the unknown." Cisco Cloud Control enters controlled availability in the United States beginning today, with broader global availability planned later.

Cisco Cloud Control: The opportunity for our partners

Customers are moving past the management of individual infrastructure domains via controller platforms. This search for seamless environments is being accelerated by agents and their workflows. To power this agentic future, we are evolving from simple disjointed visibility layers into a unified data architecture where intelligence flows seamlessly across every domain. Introducing Cisco Cloud Control Cisco Cloud Control is the unified, AI-native operations platform that brings every Cisco domain -- Networking, Security, Compute, Observability, and Collaboration -- into one environment. As a virtual front-door to all of Cisco's portfolio, it's designed for customers who need broader visibility, coherent operations, greater efficiency, and a clearer path from signal to action. Four integrated capabilities define the experience: * Cisco Cloud Control -- the platform and management plane that unifies inventory, topology, and alert management across the entire Cisco estate. For partners: the single environment from which you operate every customer engagement. * AI Canvas -- a multiplayer agentic workspace where operators and AI agents investigate issues together, across domains, in real time. For partners: where your engineers and AI agents work side-by-side on triage and resolution. * Cloud Control Studio -- a governed environment for building, customizing, and managing AI apps & agents. For partners: the toolkit for packaging your expertise into reusable, sellable IP. * AI Assistant -- natural language interface with persistent context across every connected domain. For partners: faster onboarding, faster investigation, faster answers for your customers. What this means for your business Partners reading this are asking a fair question: Does Cisco Cloud Control grow my business or threaten it? The honest answer is that it grows it -- significantly -- especially for partners who lean in early. Here's why. Today, a managed service team of 10 engineers handles roughly 50-100 customer environments. The majority of their time goes to repetitive triage, context-switching between tools, and manually correlating signals across domains. That's time your team can't bill at premium rates and capacity you can't easily scale. With Cisco Cloud Control: * AI Canvas handles first-pass triage and cross-domain correlation, freeing your engineers for the high-value investigation and strategic guidance customers actually hire you for * Workflows turn the remediation patterns you've been scripting per-customer into reusable, scalable assets * Correlated alerts surface what matters -- giving your team signal instead of noise The result: your engineers spend their time on work that demonstrates expertise, not on work that consumes it. Your capacity grows. Your margins improve. The quality of service you deliver goes up, not down. Consider what this looks like in practice. When a financial services customer hits a cross-domain incident at 2 a.m., today their partner pieces together five dashboards across networking, security, and observability before triage even begins. With Cisco Cloud Control, AI Canvas has already correlated the signals, surfaced the likely root cause, and handed the engineer a starting point -- turning a four-hour incident into a forty-minute one, and turning a stressful escalation into demonstrable expertise. "Our joint integration of Cisco AI Canvas and Cisco Cloud Control demonstrates the power of platform-level orchestration across multi-domain environments. The flexibility and API-driven architecture enabled seamless integration into our AI-native SDI Services model, accelerating operational efficiency and scalability. This reinforces the strength of our partnership and sets a strong foundation for next-generation, outcome-driven services." Dilip Kumar President & Global Head, Technology Solutions NTT DATA Inc. The expanded partner opportunity Cisco Cloud Control offers a significant shift for customers: from fragmented, domain-by-domain operations to a connected way of seeing, understanding, and acting across their Cisco environment. It also offers benefits for partners in at least two important ways -- 1) a critical role in helping customers realize unified operations, and 2) a stronger platform to help deliver and extend that value. Let's break it down more clearly in terms of customer outcome, partner roles and leveraging Cisco Cloud Control directly. 1. Serve more customers with greater consistency Cisco Cloud Control supports a consistent way to operate across connected Cisco environments, with continuity across inventory, topology, alerts, investigation, and action. For partners helping customers, that means: * guiding customers toward a unified operating model across Cisco domains * reducing fragmentation in how issues are investigated and managed * bringing consistency to how customers adopt Cisco operations tooling Aided by direct use of Cisco Cloud Control to: * investigate issues with shared context across customer environments * apply consistent service practices across engagements * increase service capacity by reducing operational friction 2. Expand from one domain to the broader Cisco estate Cisco Cloud Control gives customers a broader view across their Cisco environment, helping them understand dependencies, gaps, and opportunities beyond a single product domain. For partners helping customers, that means: * broadening the conversation from one domain to the customer's wider Cisco estate * helping customers see how networking, security, observability, compute, and collaboration relate * identifying adjacent priorities where a connected operating model improves outcomes Aided by direct use of Cisco Cloud Control to: * identify opportunities beyond the original product sale * move from point-solution conversations to wider outcome-based engagements * expand your role across more of the customer's Cisco environment over time 3. Build differentiated services around AI-assisted operations Cisco Cloud Control creates a practical path to AI-assisted operations -- grounded in operational context, governance, and real use cases. For partners helping customers, that means: * applying AI-assisted capabilities to real operational priorities * shaping use cases around industry, environment, and workflow needs * guiding adoption in a way that is usable, governed, and outcome-focused Aided by direct use of Cisco Cloud Control to: * build differentiated service offers around AI-assisted operations * use App Builder and Agent Builder within Cloud Control Studio to shape customer-specific solutions * turn repeatable expertise into higher-value consulting work 4. Build expertise now for what comes next As Cisco Cloud Control adoption grows and capabilities deepen, customers will benefit from partners who already understand how to apply the platform. For partners helping customers, that means: * guiding customers into the platform with stronger use-case alignment * helping customers sequence adoption practically * building confidence that Cisco Cloud Control can support broader operational change Aided by direct use of Cisco Cloud Control to: * build hands-on familiarity with the platform early * develop the cross-domain expertise customers will increasingly need * position your teams for deeper Cisco Cloud Control opportunities as the platform evolves 5. Connect Cisco operational insight to business workflows Cisco Cloud Control makes infrastructure insight more useful by connecting it to the workflows and systems customers already rely on. For partners helping customers, that means: * connecting operational signals to incident, change, and service workflows * translating technical insight into reporting that supports business decisions * guiding customers toward a connected operating model across infrastructure and IT operations Aided by direct use of Cisco Cloud Control to: * build higher-value consulting around workflow and platform integration * support integrations with platforms such as ServiceNow and Salesforce * expand your role from infrastructure delivery into operational and process consulting Every level of customer stands to gain For customers with one Cisco domain: Immediate AI-powered troubleshooting value -- and compounding value as they expand. AI Canvas delivers intelligent investigation in the domain they already own. For customers with two or more Cisco domains: Cross-domain intelligence from what they've already purchased. No new license. No new cost. Unified management, correlated alerts, and AI-assisted resolution across their Cisco estate. For IT leaders evaluating agentic operations: A unified, governed path to AI-driven infrastructure -- with a practical on-ramp that doesn't require rearchitecting anything. For operators: Alerts come with answers. AI agents investigate alongside the team. Context persists across shifts and handoffs. Resolution gets faster. And the messaging is simple: You already own this. Cloud Control lets you get more from what you've already invested in -- and your partner can help you unlock the full value across every domain. What partners should do now Cisco Cloud Control enters controlled availability at Cisco Live US, included with eligible Cisco product subscriptions at no additional cost. Initial availability begins in the United States, with additional regional availability to follow as data sovereignty work completes. Here's how to get involved: * Start using the platform to familiarize yourself with the capabilities and how it bridges your business with your customers' * Identify two or three target accounts where cross-domain visibility and operational coherence are already on the table * Bring Cisco Cloud Control into active conversations about simplification, resilience, and AI-assisted operations * Engage your Cisco partner team this quarter to confirm early access eligibility, enablement resources, and an adoption plan for your practice * Nominate a small internal team to build hands-on familiarity now -- the partners who develop expertise first will lead the conversations customers want to have next * Plan for Cloud Control Marketplace where you will be able to place custom AI agents tailored to your customers' verticals, create custom workflows and sell & support those agents as your own intellectual property Win-win at scale Cisco Cloud Control gives customers a unified way to operate. It gives you a stronger platform to help them do it -- and a bigger role than ever in their AI-era operations going forward. Some products or features described may be in various stages of development and offered on a when-and-if available basis.

From signal to action: The next step in Cisco AgenticOps

Every enterprise now operates at the relentless speed of digital experience. Applications are more dynamic. Users connect from everywhere. Infrastructure spans campuses, branches, clouds, service providers, and the internet. And expectations continue to rise. NetOps teams are not short on data. They are surrounded by it -- alerts, telemetry, tickets, topology changes, application signals, and user complaints arriving faster than humans can correlate them. They are being asked to resolve incidents in minutes, prevent problems before users feel them, improve performance, strengthen security, and keep the business moving across environments that change constantly. The next era of network operations cannot be built on more dashboards alone. Teams need a way to move from signal to trusted action at machine speed. That is why AgenticOps matters. It's also why agentic operations is about more than autonomy. In critical infrastructure, speed without trust creates risk. AI agents cannot operate as opaque black boxes. Operators need to understand what agents see, why they make recommendations, what actions they propose, when human approval is required, and how outcomes are verified. Every decision must be explainable, traceable, and accountable. This principle underpins Cisco Cloud Control: a unified operating model where humans and AI agents collaborate to manage critical IT infrastructure with shared context, governed action, and end-to-end accountability. This is the next step in Cisco AgenticOps -- our vision for moving complex IT workflows from manual investigation to agent-driven orchestration. The goal is not to remove operators from the loop. It is to give them an agentic workforce they can trust: agents that sense, diagnose, remediate, validate, and deploy through governance built in from the start. The future of NetOps is not human or machine. It is human-led, agent-powered operations. Across industries, senior IT leaders have told us the same thing: the network is now critical infrastructure for the business, and trust must be earned before autonomy can expand. Leaders want AI that solves real operational pain, improves user experience, increases efficiency, and gives their teams choice in how much autonomy they enable. They also want proof: reasoning, audit trails, reliability, and control. Advancing AgenticOps at Cisco Live At Cisco Live, we are advancing Cisco AgenticOps with three new capabilities that help NetOps teams move from signal to action through a governed, closed-loop workflow: * Ambient agents for networking are always-on and purpose-built to sense issues, investigate anomalies, reason over telemetry and topology, and recommend or execute actions through governed workflows. * Agentic Actions for networking gives operators one place to see, approve, audit, and control agent activity. * The Agentic Loop structures every autonomous action through five stages: sense, diagnose, remediate, validate, and deploy. It is powered by Experience Metrics, Deep Reasoning, Digital Twin, and Cisco Agentic Workflows. Together, these capabilities help teams find issues faster, understand root cause, validate the right fix, act within operator-defined controls, and verify that the user experience has recovered. Many AI tools can summarize alerts. Some automation platforms can execute predefined scripts. Cisco AgenticOps is designed to go further: to connect sensing, reasoning, workflow, validation, deployment, and experience verification in one governed operating model. That is what enterprises need before they can trust agents to act in critical environments. Agents built for network operations The starting point is the agentic workforce itself. Cisco agents for networking are designed to help NetOps teams handle the work that overwhelms them today: too many alerts, too many tools, too much context-switching, and not enough time for deep investigation. These agents monitor signals, cluster related events, identify likely causes, and propose next steps. For approved action types, they can move from recommendation to execution. For higher-risk changes, they route the decision through the right controls. Ambient agents are not waiting for a prompt. They proactively observe the environment, detect patterns, investigate anomalies, and prepare recommended actions before an operator has to ask. They are not standalone bots making isolated decisions. They operate inside Cisco Cloud Control, grounded in Cisco networking knowledge, customer telemetry, topology, configuration context, Experience Metrics, ThousandEyes network intelligence, Digital Twin validation, and governed Cisco Agentic Workflows. That context is what makes the difference between generic AI assistance and agentic operations built for enterprise networks. Agentic Actions: Where autonomy is governed If agents are going to act in the network, operators need one place to see what is happening, understand why, and govern what happens next. That is Agentic Actions for networking. Agentic Actions gives teams a clear view into agent activity across the network. Operators can see what the agent observed, what it inferred, what evidence supports the recommendation, what action it proposes, whether approval is required, what action was taken, and what outcome was verified. Every decision includes reasoning. Every action is captured in an audit trail. Every recommendation that requires oversight appears in an approvals queue. This is the difference between black-box autonomy and enterprise-ready autonomy. Operators do not lose control. They gain a governed system for deciding where agents can act independently, where approval is required, and where automation should not proceed. As these capabilities expand, customers will be able to define autonomy controls by action type, network domain, and change window. They decide where agents can act, what they are allowed to do, and when those actions can happen. This is how autonomy becomes operationally usable: not all-or-nothing, but governed, incremental, and aligned to the customer's risk model. The Agentic Loop The Agentic Loop is how agents move from signal to action in a structured, governed, and accountable way. Each stage is supported by Cisco capabilities built for trusted network operations. Sense Every loop starts with a signal. Cisco detection systems continuously watch alerts, network events, configuration changes, user experience indicators, application behavior, and other signals that something may be wrong. When an issue appears, an ambient agent begins investigating. The system does not treat every signal as a separate incident. It clusters related events, filters noise, and prioritizes what matters before routing the issue forward. For example, when 20 access points reboot simultaneously, the agent can correlate AP reboot timestamps with upstream switch reboot data using network topology. Instead of treating this as 20 separate support cases, it recognizes a single upstream switch problem and directs the fix to the right device. Experience Metrics plays a critical role in this sensing layer by turning thousands of client, device, infrastructure, and application telemetry points into simple, actionable indicators of user experience. Instead of forcing teams to interpret every raw signal, it helps them understand whether users are having a good experience across wired, wireless, and application environments. By combining application-aware traffic inspection built into wireless access points and switches with proactive synthetic measurements, customers gain real-time insight into how users experience applications across the network. That holistic view helps agents detect abnormalities earlier, understand impact more clearly, and begin root cause analysis with the user experience in mind. Diagnose Once the signal is understood, the agent diagnoses the issue using telemetry, topology, configuration context, and Cisco networking expertise. Many problems can be resolved quickly with a focused answer. The harder issues -- ambiguous symptoms, cross-domain dependencies, or conditions outside a standard playbook -- require deeper investigation. That is where Deep Reasoning comes in. Deep Reasoning is designed for the problems that simple playbooks cannot solve. It uses a self-correcting diagnostic process grounded in a Cisco-authored library of standardized networking skills. These skills span wired, wireless, WAN, and security domains, encoding the diagnostic methodology and decision logic that an experienced network engineer would apply. That grounding matters. In production networks, a confident wrong answer is not good enough. Imagine an agent that diagnoses a Layer 2 loop and recommends shutting down a port, even though the topology is linear and there is no loop at all. That confident-but-wrong answer is what hallucinated AI looks like in production, and it is exactly the kind of mistake that erodes trust in autonomy. With Deep Reasoning, the agent checks its work against observable network signals such as topology, port state, and MAC behavior. The agent reasons. The network confirms. In early deployments, this approach is already surfacing issues traditional monitoring can miss -- from unstable backup cellular WAN links across multiple sites to access points falling into mesh repeater mode because an unmanaged upstream switch had gone offline. Deep Reasoning Mode is also available in AI Assistant, giving operators investigative responses rather than surface-level summaries. It brings Cisco expert-informed diagnostics, in natural language, to every operator on the team. After a Deep Reasoning session, an information security leader at a state public-sector agency told us: "This is a huge leap forward. It's giving me another member of my team." These are the kinds of problems AgenticOps is designed to find: real issues, in real production networks, before they become larger incidents. Remediate Once the issue is diagnosed, the agent generates the proposed fix through Cisco Agentic Workflows. Remediation does not happen through improvised API calls or one-off scripts. It runs through validated automation, with audit logging and step-by-step traceability. The proposed remediation is visible in Agentic Actions, where operators can review the recommendation and decide what happens next. If the action is pre-approved, the agent can continue. If it requires oversight, it moves into the approval queue. This is where AgenticOps begins to change the operating model. The agent is not simply telling the operator what might be wrong. It is helping prepare the next best action, presenting the evidence, and routing the action through the right controls. Validate Before a change is deployed, the system validates the action based on risk. Some actions may be low-risk and pre-approved. Others require deeper validation. For changes such as topology modifications, security policy updates, or large-scale configuration changes, the proposed fix can be tested against a Digital Twin before it touches production. Cisco Digital Twin gives teams a virtual environment to model network behavior and evaluate proposed changes before they reach production. For higher-risk actions, agents can test expected impact, assess blast radius, and identify potential unintended consequences before deployment. That simulation helps answer the questions operators care about most: Will this fix the issue? What is the potential blast radius? Could the change create unintended consequences? Is it safe to deploy? Not every action needs a full simulation. The system applies the right level of validation based on risk and blast radius. That is what makes autonomy practical in the real world. It allows teams to move faster where the risk is low and apply deeper assurance where the stakes are higher. Deploy Once validated, the action is deployed according to the customer's controls. Some actions can be executed autonomously. Others require human approval. The choice depends on the action type, the network, the change window, and the policies the operator has defined. But deployment is not the end of the loop. It is the moment the system proves whether the action worked. After a change is deployed, Cisco uses ThousandEyes synthetic testing and Experience Metrics to help verify that the user experience has recovered. The loop does not close simply because infrastructure reports green. It closes when the outcome is confirmed. The goal is not just to clear an alert. The goal is to restore the experience. Earning trust, one decision at a time Human-only NetOps can no longer keep pace with the complexity and speed of modern digital environments. But enterprises will not trust critical infrastructure to black-box autonomy. That is why Cisco AgenticOps is built around a different principle: autonomy must earn trust every time it acts. It earns trust by sharing context. By explaining its reasoning. By operating through governed workflows. By validating higher-risk changes before they touch production. By giving operators control over where agents can act and where approval is required. And by closing the loop only when the user experience is restored. This is the path to trusted autonomous network operations: human-led, agent-powered, and outcome-validated. It is also a journey. With every advance in AI reasoning, every new skill added to the library, and every customer deployment that sharpens our models, the scope of what Cisco agents can do autonomously will grow. But the principle will not change. Always governed. Always accountable. Always earning trust, one decision at a time. Learn more about Cisco AgenticOps. See more Enterprise Networking news from Cisco Live Las Vegas 2026: The Agentic Workplace Runs on Cisco by Anurag Dhingra Trust at Machine Speed: Building Secure Campus Networks for the AI Era by Michael Dickman Cisco Unveils Multicloud Fabric in Cloud Control: Network Ready for the AI Era by Rohit Agarwalla

Cisco Unveils Cloud Control, AI Canvas, Quantum Security Push At Cisco Live

Cisco rolls out new agentic AI, security and quantum-readiness technologies at Cisco Live, centering on Cisco Cloud Control, AI Canvas and Agent Studio to help customers manage AI-driven infrastructure, improve operations and telemetry, and prepare networks and encryption for post-quantum threats. Cisco Tuesday morning kicked off its annual Cisco Live conference in Las Vegas with the introduction of several new technologies aimed at bringing agentic AI and improved security to data center and cloud infrastructure. Cisco's full-court press on agentic AI and security is centered in large part around Cisco Cloud Control, a unified platform spanning the company's networking, security and observability portfolios, which was originally previewed at Cisco Live 2025. Jeff Schultz, senior vice president of portfolio strategy for the San Jose, Calif.-based company's product organization, said at a pre-Cisco Live press conference that the new agentic AI push comes as the industry starts to view AI agents as co-workers to human employees. [Related: Cisco Eyes Astrix Security To Lock Down AI Agents In Potential $350M Deal: Report] "What we are really seeing now as our customers deploy agent AIs is that literally every action that an agent takes is a combination of both a routing challenge, a trust decision and a telemetry event," Schultz said. "Because they are deployed throughout the entire enterprise -- in data centers, as well as in campuses and branches -- every action they take is driving network traffic and network decisions. Security is a critical issue, and telemetry is now available with everything that they do to really help us understand their actions and their outcomes." While the ChatGPT world resulted in "spiky" traffic and demands on data center and cloud infrastructure, he said, agentic AI is creating a long, sustained demand on infrastructure. And agents collaborating with other agents and accessing systems introduce a lot of risk to infrastructure, Schultz added. "We are really very firmly focused on being this critical infrastructure layer for the AI era," he said. Under the One Cisco strategy, Cisco looks at networking, security, observability and collaboration technologies across a couple of key outcomes, Schultz said. * AI-ready data centers housing new digital workers that are not limited to individual buildings or networks of buildings, but extend to the edge, campus, and branch. * Future-proof workplaces where, as the workplace changes and as agents, robots and other physical AI come into play, infrastructure and security perspectives change. * The need for secure global connectivity in the work done around service providers and telecoms because data centers and workplaces don't live in isolation. "We are innovating significantly in ways to connect these data centers together in high-performance ways and connect data centers to workplaces," he said. "And digital resilience is a foundation. So being able to leverage data in new and unique ways to understand anything that can put your business at risk, whether it's an infrastructure issue, whether it's a security threat, or whether it's agents that are working outside of their guardrails and are working outside of their rules, all of this is accelerated by Cisco AI." Cisco Cloud Control The latest version of Cisco Cloud Control was born from the question of what business should look like when humans and agents start to work together, said DJ Sampath, senior vice president and general manager of Cisco's AI Software and Platform Group. "It's not just about humans clicking through dashboards, trying to keep up, but a true collaborative operating model where agents are doing the heavy lifting and humans are staying in control of what matters," Sampath said. The focus of Cisco Cloud Control is on agentic operations, or AgenticOps, Sampath said. "Cisco Cloud Control is a platform where all of this comes to life," he said. "We're focusing on enabling cross-domain telemetry with the power of the purpose-built models that we have built with over 40 years of Cisco data and knowledge and together creating trusted agents to serve as digital co-workers." Cisco Cloud Control, built on the Cisco Data Fabric powered by the Splunk Platform as its unified data layer, provides cross-domain telemetry for all of a customer's domains across all Cisco products, platforms, services and agents via what Sampath said were purposeful models aimed at solving specific types of tasks and issues. These include a deep network model, a cybersecurity model called Foundation-Sec, and frontier models. "It's really about matching the right intelligence to the right type of task, and that's what we're doing here," he said. "We use a combination of the telemetry and the purposeful models to be able to build these trusted agents. When you combine the model intelligence with the telemetry, that's really what makes these agents drive the outcomes that you're looking for." Cisco has a strong end-to-end play that works only when utilizing all the components, said Lane Irvine, alliance leader at Long View Systems, a Calgary, Alberta-based solution provider and longtime Cisco channel partner. "Now we can bring that together to take advantage of the Cisco platform," he said. "It's a great opportunity for them. I think it's an important move as we start leading into Cisco President and Chief Product Officer Jeetu Patel taking over all the platforms. They are trying to create a more synergistic environment for that joint development, and that's where I think Cloud Control will be a strong move." With Cisco Cloud Control, Cisco wants everything to be managed from a central control plane, said Faisal Bhutto, president and CEO of Alykas, a Houston-based solution provider and longtime Cisco channel partner. The reality is that OEMs like Cisco know their tech stacks better than anybody, Bhutto told CRN. "Cisco knows Meraki, how it works, how it operates," he said. "Cisco knows the Catalyst line, UCS, the whole fabric and how they operate. So customers, instead of creating their own automation tools to run a multisite WAN or SD-WAN or a big SaaS deployment, they want OEMs like Cisco to come up with agentic AI they can use along with their human resources. The promise is there. I can see that becoming more and more real, where our customers can be in Cloud Control with access to agentic bots that take action, pending approval." Customers won't be comfortable letting agentic AI take actions, Bhutto said. "But there definitely is a huge value where customers can say, 'Instead of me having all these NOC [Network Operations Center) resources, I don't really need them because that's what agentic AI can handle, but I will keep decision-making in my control,'" he said. "Should there be a config change that has to be rolled out or some other modification, go into Cloud Control and approve that change for it to get pushed out. I think customers are going to be more comfortable with that." Cisco Cloud Control Tuesday entered controlled availability in the U.S. Beyond Cisco Cloud Control Within Cisco Cloud Control lies Cisco AI Canvas, an environment for humans and AI agents to work together in real time to investigate, correlate and resolve complex issues, Sampath said. AI Canvas is also going into controlled availability, he said. To make it work, Cisco Tuesday also introduced Agent Studio, Sampath said. "Agent Studio allows you to seamlessly bring in third-party agents that help you, and we also additionally on top of that help you design your own agents and consume the agents that we at Cisco are building for you," he said. "[Agent Studio] systematically brings all of these pieces together, but we also recognize that every single IT administrator, security operator, every single operator, is turning into a builder today." Using Codex and Claude Code, businesses can build complex systems quickly, Sampath said. "We want to help teams to customize each and every single operation in a way that's unique to their workflows, and so we're launching an app builder, which is a space for creating and customizing applications and workflows inside of Cisco Cloud Control, which essentially makes Cloud Control a harness that is powered by Open AI's Codex that is natively integrated into Cisco Cloud Control," he said. Cisco has made a lot of investments into AI and agentic AI, Long View Systems' Irvine said. "As customers look at their environments now, they've got higher expectations of what they need to get from that environment, and that's where Cisco and others really need to use agentic AI agents to go," he said. "The question is, 'How are we going to manage this from an operational perspective? How are we going to get insight into what's going on?' We want to be able to get predictive insight. ... Agentic AI really opens the door for that." Cisco's work around AI Canvas is also interesting, Irvine said. "When I'm having a network issue, it might be my network is slow," he said. "What does that mean? Where is that impacted? I need to be able to look at everything from security to infrastructure to wireless to end-user information. I need to look at that entire stream to analyze and quickly identify where there's an issue, even the application level. And with the Cisco portfolio, we'll have a much better insight into that. AI Canvas provides that platform to do it across the infrastructure." Cisco has also been busy on the quantum computing front, said Anurag Dhingra, senior vice president and general manager of Cisco's Enterprise Connectivity and Collaboration Group. Cisco has been advancing its post-quantum cryptography, and earlier this year introduced post-quantum crypto support for its smart switches and routers, Dhingra said. "Now we're extending that to SD-WAN links to protect all traffic in transit," he said. Cisco this week also unveiled significant enhancements to its post-quantum encryption capabilities, including signed digital images of the software-hardware trust angle secure boot and making its crypto libraries ready for a post-quantum world, Dhingra said. Customers are certainly starting to look at what it takes to be quantum-ready, Irvine said. "I don't know that everybody fully appreciates the risk, but it is something that Cisco absolutely needs to lean into," he said. "We need to educate our customers about quantum-ready and the requirements for quantum-ready infrastructure because the 'harvest now, decrypt later' move is already happening. We have to be ready for post-quantum cryptography." In addition, Cisco unveiled new resilient infrastructure services to help customers and partners integrate the company's new security technologies, said Bhaskar Jayakrishnan, senior vice president of engineering for customer experience at Cisco. Resilient infrastructure services include a three-step approach, he said: assess exposure, modernize infrastructure and build defensive resilience. More self-healing, proactive and preventative protection helps employees and customers worry less and focus more on speed and execution instead of protecting themselves from the bad guys, Alykas' Bhutto said. "The reality is the same kind of capabilities over time do get into the hands of bad guys as well, and they constantly find ways to penetrate and exploit," he said. "Why do I have a bad feeling that an LLM like a Mythos might exist somewhere else that somebody may be using already to exploit security issues? Remember, in security, the name of the game is not the threat actors or your private threat actor. There are state actors involved, and that's a different legal thing. Who is to say that an enemy state is not already figuring out how to get ahead of it." Also new from Cisco is its Quantum Ready Assessment as part of Cisco IQ, which works with new quantum-ready hardware to identify areas most at risk for "harvest now, decrypt later" attacks, Jayakrishnan said. "What we help our customers with is to evaluate their landscape so that they know what to prepare for which network devices to either replace or upgrade to get to quantum readiness," he said. The result will be quantum-safe assessments, Jayakrishnan said. "We give our customers a top-down and a bottom-up view of whether they can take a look at their overall estate, compare how ready they are according to the regulatory requirements of their own region, and what are the next set of actions, whether they need to upgrade the hardware, upgrade software, take care of features, or how much of their estate is quantum safe," he said. Quantum readiness is less about future technology and more about how to modernize and manage existing encryption, identity and trust in enterprises, Irvine said. "Without that encryption and security in today's environment, whatever's coming is irrelevant, and we've got to make sure that we're prepared for it now even if the immediate impacts are limited," he said. Quantum resilience has been a topic for years, although AI has kind of overshadowed it in the past couple of years, Bhutto said. "Everybody has to figure out and ensure that as quantum computing becomes more readily available and accessible at certain price points, the encryption mechanisms and keys we use today are quantum-resilient," he said. "Otherwise, they can easily be broken. Customers worry about it, but to be honest, the AI talk track has almost overshadowed everything else. Quantum security used to be a big topic two years ago. But now customers are so focused on agentic AI that they're not talking about it as much."

The path to autonomous networking: Meet the agents helping win the AIOps race

Imagine managing a racing team during a championship race. Success is not determined by a single driver alone. It depends on a highly coordinated team operating in real time. Engineers analyze telemetry continuously. Strategists predict changing track conditions. Pit crews execute precision adjustments. Analysts identify risks before they become failures. Every team member specializes in a different function, yet all must work together dynamically to keep the car performing at peak efficiency under constantly changing conditions. Modern network operations face a very similar challenge. Today's large-scale infrastructure networks are living, constantly evolving systems. Traffic patterns shift unpredictably. AI is reshaping workloads and infrastructure demands. Multivendor environments generate overwhelming volumes of telemetry while operational data remains siloed across domains. Even with automation in place, operations teams often remain trapped in reactive workflows, manually troubleshooting incidents, correlating alarms across disconnected domains, and resolving issues only after services are impacted. It's no longer enough to simply automate individual tasks. Instead, teams must enable intelligent operational teamwork at machine speed. This is where Cisco Crosswork Network Automation is changing the paradigm of network operations. At Cisco Live, we are introducing the next evolution of the Cisco Crosswork Network Automation suite: Cisco Crosswork® AI, a secure, scalable multi-agentic framework designed to act as an extension of your network team. Much like a championship racing team, specialized AI agents work together continuously, reasoning through problems, identifying risks, troubleshooting complex issues, validating operational intent, and recommending corrective actions in real time. Cisco Crosswork Network Automation is the heartbeat of our Agile Services Networking architecture for AI connectivity, enabling teams to operate resilient, AI-ready networks at scale. Empower your team with a secure, scalable multi-agentic platform Cisco Crosswork AI is a transformative multi-agentic AI framework integrated within Cisco Crosswork Network Automation, debuting with a powerful team of specialized agents designed to tackle the most demanding operational challenges. Additionally, Crosswork AI will be integrated into Cisco AI Canvas (the generative user interface that's part of Cisco Cloud Control), to provide unified, cross-domain management. * Deep network troubleshooting: Like race engineers diagnosing performance anomalies in real time, these agents reason through complex issues, evaluate hypotheses, and pinpoint root causes with precision to reduce mean time to repair (MTTR), minimizing service downtime. * Risk factor detection: Similar to strategists anticipating tire degradation or changing track conditions before they impact race performance, these agents uncover hidden combinations of variables that trigger network degradation, helping resolve issues proactively before services are affected. * Configuration drift remediation: Much like the team members who help ensure every race car component remains calibrated to exact specifications throughout the race, these agents continuously detect configuration drift, identify variances using AI trained on device configurations, and provide remediation recommendations to maintain operational consistency. This extensible framework supports an expanding ecosystem of agents, both out-of-the-box and customized, with the ability to bring your own agents through the agent development kit (ADK). With capabilities such as agent evaluation, a knowledge graph, and an agent catalog, the framework establishes a strong operational foundation for agentic operations, enabling networks to evolve from automated systems to intelligent operational teams. The strength of the Crosswork AI multi-agentic architecture has already been recognized on the industry stage. The platform was featured in the TM Forum Catalyst project "Game X," selected as a Moonshot project, an honor reserved for initiatives demonstrating transformative potential for the telecommunications industry. This recognition underscores how Crosswork AI can power next-generation, AI-driven network operations at scale. Automate and assure the complete lifecycle of network operations In racing, winning is not about excelling in a single lap. Success depends on orchestrating the entire race lifecycle, from pre-race planning and strategy simulation to live telemetry analysis, pit-stop execution, and continuous optimization throughout the race. The same principle applies to modern network operations. The true value of Crosswork lies in its ability to manage the entire network lifecycle, from device onboarding to end-to-end service delivery including planning, orchestration, assurance, optimization, and troubleshooting. Crosswork enables organizations to: * Unify visibility: Similar to how race engineers maintain complete visibility into vehicle performance, tire conditions, fuel strategy, and competitor positioning, Crosswork provides comprehensive end-to-end visibility across topology, telemetry, and service health with AI-enhanced operational clarity. * Correlate insights: Just as racing teams transform thousands of telemetry signals into strategic race decisions, Crosswork converts multisource telemetry, service, and state data into actionable intelligence that improves operational precision and business outcomes. * Drive action: Like a smooth, perfectly timed pit stop that restores peak performance without disrupting the race, Crosswork enables effective, streamlined lifecycle operations through automated provisioning, AI-guided remediation, and closed-loop operational workflows. The platform continuously translates business and operational intent into network reality. Enable automation that's multivendor, open, and extensible The modern network is rarely a single-vendor environment. Crosswork leverages standard APIs and protocols, including NETCONF, RESTCONF, TAPI, OpenConfig, and native YANG data models, as well as streaming telemetry from gRPC, BGP-LS, and gNMI, to enable effective automation across multivendor environments. By adhering to open, industry-recognized interfaces, Crosswork allows for interoperability without vendor lock-in, enabling teams to easily adopt new technologies, platforms, and domains. This commitment to openness is backed by independent, real-world validation across top-tier service provider, public sector, and enterprise multivendor networks across the world. In the European Advanced Networking Test Center (EANTC) interoperability showcase, Crosswork successfully automated services across a diverse ecosystem of vendor devices, demonstrating multivendor richness in a live, heterogeneous environment. We recognize that real-world networks often include legacy systems and specialized devices that do not support standardized interfaces. To solve for multivendor automation in any environment, Crosswork provides an open SDK for customized packages and agents to handle proprietary integrations, helping ensure no device or domain is left behind. Whether your environment is built on the latest standards-based platforms or includes a mix of legacy and proprietary systems, Crosswork delivers consistent, end-to-end automation, protecting existing investments while accelerating modernization. Accelerate your autonomous networking journey This multi-agentic evolution is more than just a feature update. It is a fundamental shift in how network assurance and automation are approached. We look forward to showcasing these innovations at Cisco Live 2026 in Las Vegas and demonstrating how Cisco Crosswork AI, part of our broader Cisco vision for agentic operations, can help organizations accelerate their journey toward resilient, intelligent, and autonomous networking. Visit us at the Secure Global Connectivity area inside the World of Solutions to experience the multi-agentic framework in action and discover how Cisco Crosswork AI can help you operate with the precision, intelligence, and coordination of a championship-caliber network operations team.

Cisco Unveils Cloud Control Platform to Power Agentic AI for Critical IT Infrastructure

Cisco Cloud Control is the single management plane that brings a customer's entire estate into one environment -- one login, one view. It's a new way to run critical infrastructure that ties together: Cross-domain telemetry. The rich data flowing across networking, security, observability, collaboration, and more - comes together in Cloud Control so humans and agents can act on the same information to address key business imperatives like uptime, agent behavior, and tokenomics. Purpose-built models. Cloud Control reasons across complex problems with the right mix of purpose-built and frontier models -- including Cisco's Deep Network Model, grounded in 40 years of Cisco operational networking data. The result is system intelligence that scales with the complexity of the problem, not the size of the model alone. Trusted agents. Through Cisco Cloud Control, operators will be able to work with autonomous agents that can follow a structured path from signal to action: spotting trouble, identifying causes, carrying out fixes, testing changes before deployment, and confirming the user experience has recovered. These agents will be powered by Cisco telemetry and purpose-built models, and they will leverage other capabilities, such as Expanded Experience Metrics, Deep Reasoning, Digital Twin, and Cisco Agentic Workflows. Teams will be able to automate network ops with an agentic loop, while keeping actions visible and governed. Cisco AI Canvas. A multiplayer, generative workspace where operators and agents work from the same live evidence to investigate and resolve complex issues together in real time. Context persists across shifts and escalations, so nothing is lost, and nothing is repeated. Cloud Control Studio. The design space that unlocks two customization environments. Agent Builder lets customers build agents for Cloud Control tailored to their own policies and workflows, with the ability to connect to more than 50+ third-party platforms and tools through native connectors or the open Model Context Protocol (MCP). App Builder lets customers build and publish apps and workflows for Cloud Control from natural-language prompt, with OpenAI Codex, an agentic platform that helps you build and ship with AI, built in. Everything built in Studio -- plus agents and apps from across Cisco's ecosystem -- can be published to Cloud Control Marketplace.

Cisco's New Weapon Against AI-Accelerated Cyber Attacks

Cisco Cloud Control brings together human operators and trusted AI agents to run the world's most critical systems In an agentic AI world, organizations must act and defend at machine speed and scale. Cisco Cloud Control, unveiled today at Cisco Live, is a unified platform built for humans and AI agents to manage, monitor and defend critical IT infrastructure -- and the foundation for Cisco's AgenticOps operating model. With one login, Cisco Cloud Control delivers a single view of Cisco networking, security, compute, observability, and collaboration in one secure environment. People and agents work from a single data layer, sharing the same operational context and the same system of action, while humans stay in control. Customers can build their own applications and agents using natural language directly within the platform, which also connects to a large ecosystem, including AWS, Linear, Microsoft, PagerDuty, ServiceNow, Slack, and Google Cloud, which now includes Wiz. "AI agents reason and act continuously at software speed, and that changes everything about how we scale, manage, and defend our critical infrastructure," said Jeetu Patel, President and Chief Product Officer, Cisco. "Cisco Cloud Control is a command center for agentic AI: a platform where your team and your AI agents work together, in the same environment, with the same information, and with humans in control." One platform for humans and agents to run the agentic enterprise Cisco Cloud Control is the single management plane that brings a customer's entire estate into one environment -- one login, one view. It's a new way to run critical infrastructure that ties together: * Cross-domain telemetry. The rich data flowing across networking, security, observability, collaboration, and more - comes together in Cloud Control so humans and agents can act on the same information to address key business imperatives like uptime, agent behavior, and tokenomics. * Purpose-built models. Cloud Control reasons across complex problems with the right mix of purpose-built and frontier models -- including Cisco's Deep Network Model, grounded in 40 years of Cisco operational networking data. The result is system intelligence that scales with the complexity of the problem, not the size of the model alone. * Trusted agents. Through Cisco Cloud Control, operators will be able to work with autonomous agents that can follow a structured path from signal to action: spotting trouble, identifying causes, carrying out fixes, testing changes before deployment, and confirming the user experience has recovered. These agents will be powered by Cisco telemetry and purpose-built models, and they will leverage other capabilities, such as Expanded Experience Metrics, Deep Reasoning, Digital Twin, and Cisco Agentic Workflows. Teams will be able to automate network ops with an agentic loop, while keeping actions visible and governed. * Cisco AI Canvas. A multiplayer, generative workspace where operators and agents work from the same live evidence to investigate and resolve complex issues together in real time. Context persists across shifts and escalations, so nothing is lost, and nothing is repeated. * Cloud Control Studio. The design space that unlocks two customization environments. Agent Builder lets customers build agents for Cloud Control tailored to their own policies and workflows, with the ability to connect to more than 50+ third-party platforms and tools through native connectors or the open Model Context Protocol (MCP). App Builder lets customers build and publish apps and workflows for Cloud Control from natural-language prompt, with OpenAI Codex, an agentic platform that helps you build and ship with AI, built in. Everything built in Studio -- plus agents and apps from across Cisco's ecosystem -- can be published to Cloud Control Marketplace. Cisco Cloud Control enters Controlled Availability in United States today, with Global Availability to follow. Security for the Mythos era, fused directly into the infrastructure Reactive defense is no longer enough when the window between vulnerability and exploit has collapsed from weeks to minutes. As a charter member of Anthropic's Project Glasswing and OpenAI's Daybreak, Cisco stress-tests its own products using the latest frontier AI models -- finding the weaknesses adversaries would, before they can. Cisco isn't keeping that advantage to itself: through the recently open-sourced Foundry Security Spec, every defender can apply that same rigor to their AI-driven security evaluations. Cisco is expanding protections across its infrastructure to shield customers from new vulnerabilities quickly following discovery -- with Cisco Cloud Control as the security command center where defense plays out in real time. Always-on defense across the infrastructure Live Protect acts as a digital immune system for Cisco products, shielding them from newly discovered and prioritized vulnerabilities for supported platforms at runtime -- no reboots, no upgrades, no maintenance windows. Now available in N9000 series switches and included with the Nexus One product entitlement, Live Protect is expanding to more products in the Cisco Portfolio in the coming months, starting with campus and branch smart switches, followed by secure routers later in the year. Hybrid Mesh Firewall extends unified protection across networks, applications, and Cisco and third-party firewalls -- limiting the blast radius when something goes wrong. Protecting agents from the world, and the world from agents AI agents are joining the workforce, working alongside humans, and taking on tasks that require a secure environment. At RSAC, Cisco announced an array of new innovations to protect agents from the world, protect the world from agents, and detect and respond to issues at machine speed (link to RSAC blog). Today, Cisco announced further enhancements across its agentic security offerings, from AI Defense, to Zero Trust for agents, to the Agentic SOC. Clear path to quantum-safe infrastructure. "Harvest now, decrypt later" attacks are already happening, collecting encrypted data to unlock when quantum capabilities catch up. Cisco is turning tomorrow's threat into a plan enterprises can start building today: * New Quantum-safe communications advancements across Cisco's core portfolio. With a commitment to enable quantum-safe communications capabilities across the majority of Cisco's core portfolio by December 2026, Cisco is extending post-quantum protection to the systems where the most sensitive enterprise traffic flows. * Quantum-safe by default for new infrastructure. Starting today and moving forward, any newly introduced campus, branch and data center routers, switches, and firewall series will launch with quantum-safe secure boot. This builds on the same quantum-safe technology already shipping in campus smart switches. * The new Quantum Ready Assessments, available through Cisco IQ, identify the assets most exposed to "harvest now, decrypt later" attacks -- and where to start. Global availability planned for July 2026. * The new Quantum Resilience Framework gives enterprises a structured approach to post-quantum cryptography across two pillars: quantum-safe communications and quantum-safe products. Long-term resilience with Cisco To help customers navigate this new era, Cisco Services is announcing new capabilities:

Cisco Unveils Multicloud Fabric in Cloud Control: Network Ready for the AI Era

Enterprise applications no longer live in one cloud. Today, a single workload can span compute, data, and AI services across multiple cloud providers -- while users across thousands of branches and campuses need to access applications wherever they run, and those applications need to communicate across clouds. The underlying network has not kept pace. Traffic funnels through centralized hubs before reaching the cloud-hosted enterprise application. Each cloud runs its own networking stack with no common management layer. Stretched IT teams are left navigating fragmented operations across clouds; centralized cloud connectivity architectures can add latency, failure points, and scale limits; and bolted-on security can increase risk. AI raises the bar further. According to the AI Impact on Wide Area Networks report from Cisco, agentic AI workflows generate approximately 450% more network traffic than the same tasks performed manually -- with roughly 70% of that traffic being AI inference that chains across large language models (LLMs), SaaS platforms, and private data sources that could be hosted in different clouds. Each cloud boundary can create a visibility gap. These workflows are latency-sensitive, multi-step, and span providers by design -- yet many enterprises lack a unified way to see, secure, or troubleshoot the entire path. A simpler, more scalable approach Today at Cisco Live 2026, we're introducing Cisco Multicloud Fabric -- a multicloud network-as-a-service offering available through Cisco Cloud Control. It delivers a single fabric for secure site-to-cloud and cloud-to-cloud networking that scales on demand and is operated by Cisco. Three things set it apart: One unified platform. Manage site-to-cloud and cloud-to-cloud connectivity from Cisco Cloud Control. Onboard sites and cloud environments, define intent-based connectivity, set security policies per connection, and monitor performance -- all in one place. An on-demand global fabric. Cisco deploys and operates virtual points of presence (vPoPs) across cloud providers and cloud regions. Whether it's a new workload in an existing cloud or a site that needs to reach a new environment, connectivity is ready when you are. Security and visibility built in. Zero Trust routing helps ensure nothing connects by default -- every connection must be explicitly defined. Security policy follows the traffic, with cloud firewall service chaining enforceable per connection. Cisco ThousandEyes agents embedded in each vPoP provide detailed monitoring from branch to cloud as well as cloud to cloud. Today, customers with Cisco SD-WAN deployments, starting with Cisco Meraki MX, can connect across Amazon Web Services (AWS), Microsoft Azure, and Google Cloud using Cisco Multicloud Fabric. Applications in brownfield cloud networking deployments can be migrated onto the fabric at your own pace. "We are evolving our network architecture to better support hybrid and multicloud environments. Cisco Multicloud Fabric will enable us to establish consistent connectivity, centralized policy control, and improved end-to-end visibility across clouds and on-premises infrastructure. This helps reduce operational complexity while maintaining the performance and reliability required for our business." -- Carsten Breuer, Pfeifer & Langen Extending the AI-ready secure network Cisco Multicloud Fabric extends our strategy for building networks that are simpler to operate, secure by design, and ready for AI. It brings operational simplicity and AgenticOps capabilities to multicloud environments, with AI-driven insights that help teams detect, diagnose, and resolve issues across clouds with the same speed and confidence they have with campus and branch today. And because Multicloud Fabric is delivered as a service, Cisco deploys and operates the vPoPs across cloud environments -- so teams do not have to. It provides capabilities to natively integrate security into the connectivity fabric. As AI agents begin to operate across clouds, the network is no longer a passive transport layer -- it becomes part of the intelligence stack. Cisco Multicloud Fabric is built for that reality. "As a Cisco partner, we're excited about the potential of Multicloud Fabric to solve real connectivity challenges our enterprise customers face. It will provide an easy way to control site-to-cloud and east-west traffic across multiple clouds and organizations, something our customers have struggled with for years." -- Matthias Wülfert, Computacenter Cisco Multicloud Fabric At a Glance Additional resources Cisco Multicloud Fabric solution page Cisco Multicloud Fabric Solution Overview See more Enterprise Networking news from Cisco Live Las Vegas 2026: The Agentic Workplace Runs on Cisco by Anurag Dhingra Trust at Machine Speed: Building Secure Campus Networks for the AI Era by Michael Dickman Cisco Unveils Multicloud Fabric in Cloud Control: Network Ready for the AI Era by Rohit Agarwalla

Cisco Launches Agentic Platform to Defend and Automate Critical IT Infrastructure

Cisco Cloud Control brings together human operators and trusted AI agents to run the world's most critical systems In an agentic AI world, organizations must act and defend at machine speed and scale. Cisco Cloud Control, unveiled today at Cisco Live, is a unified platform built for humans and AI agents to manage, monitor and defend critical IT infrastructure -- and the foundation for Cisco's AgenticOps operating model. With one login, Cisco Cloud Control delivers a single view of Cisco networking, security, compute, observability, and collaboration in one secure environment. People and agents work from a single data layer, sharing the same operational context and the same system of action, while humans stay in control. Customers can build their own applications and agents using natural language directly within the platform, which also connects to a large ecosystem, including AWS, Linear, Microsoft, PagerDuty, ServiceNow, Slack, and Google Cloud, which now includes Wiz. "AI agents reason and act continuously at software speed, and that changes everything about how we scale, manage, and defend our critical infrastructure," said Jeetu Patel, President and Chief Product Officer, Cisco. "Cisco Cloud Control is a command center for agentic AI: a platform where your team and your AI agents work together, in the same environment, with the same information, and with humans in control." One platform for humans and agents to run the agentic enterprise Cisco Cloud Controlis the single management plane that brings a customer's entire estate into one environment -- one login, one view. It's a new way to run critical infrastructure that ties together: * Cross-domain telemetry. The rich data flowing across networking, security, observability, collaboration, and more - comes together in Cloud Control so humans and agents can act on the same information to address key business imperatives like uptime, agent behavior, and tokenomics. * Purpose-built models. Cloud Control reasons across complex problems with the right mix of purpose-built and frontier models -- including Cisco's Deep Network Model, grounded in 40 years of Cisco operational networking data. The result is system intelligence that scales with the complexity of the problem, not the size of the model alone. * Trusted agents. Through Cisco Cloud Control, operators will be able to work with autonomous agents that can follow a structured path from signal to action: spotting trouble, identifying causes, carrying out fixes, testing changes before deployment, and confirming the user experience has recovered. These agents will be powered by Cisco telemetry and purpose-built models, and they will leverage other capabilities, such as Expanded Experience Metrics, Deep Reasoning, Digital Twin, and Cisco Agentic Workflows. Teams will be able to automate network ops with an agentic loop, while keeping actions visible and governed. * Cisco AI Canvas. A multiplayer, generative workspace where operators and agents work from the same live evidence to investigate and resolve complex issues together in real time. Context persists across shifts and escalations, so nothing is lost, and nothing is repeated. * Cloud Control Studio. The design space that unlocks two customization environments. Agent Builder lets customers build agents for Cloud Control tailored to their own policies and workflows, with the ability to connect to more than 50+ third-party platforms and tools through native connectors or the open Model Context Protocol (MCP). App Builder lets customers build and publish apps and workflows for Cloud Control from natural-language prompt, with OpenAI Codex, an agentic platform that helps you build and ship with AI, built in. Everything built in Studio -- plus agents and apps from across Cisco's ecosystem -- can be published to Cloud Control Marketplace. Cisco Cloud Control enters Controlled Availability in United States today, with Global Availability to follow. Security for the Mythos era, fused directly into the infrastructure Reactive defense is no longer enough when the window between vulnerability and exploit has collapsed from weeks to minutes. As a charter member of Anthropic's Project Glasswing and OpenAI's Daybreak, Cisco stress-tests its own products using the latest frontier AI models -- finding the weaknesses adversaries would, before they can. Cisco isn't keeping that advantage to itself: through the recently open-sourced Foundry Security Spec, every defender can apply that same rigor to their AI-driven security evaluations. Cisco is expanding protections across its infrastructure to shield customers from new vulnerabilities quickly following discovery -- with Cisco Cloud Control as the security command center where defense plays out in real time. Always-on defense across the infrastructure Live Protect acts as a digital immune system for Cisco products, shielding them from newly discovered and prioritized vulnerabilities for supported platforms at runtime -- no reboots, no upgrades, no maintenance windows. Now available in N9000 series switches and included with the Nexus One product entitlement, Live Protect is expanding to more products in the Cisco Portfolio in the coming months, starting with campus and branch smart switches, followed by secure routers later in the year. Hybrid Mesh Firewall extends unified protection across networks, applications, and Cisco and third-party firewalls -- limiting the blast radius when something goes wrong. Protecting agents from the world, and the world from agents AI agents are joining the workforce, working alongside humans, and taking on tasks that require a secure environment. At RSAC, Cisco announced an array of new innovations to protect agents from the world, protect the world from agents, and detect and respond to issues at machine speed (link to RSAC blog). Today, Cisco announced further enhancements across its agentic security offerings, from AI Defense, to Zero Trust for agents, to the Agentic SOC. Clear path to quantum-safe infrastructure. "Harvest now, decrypt later" attacks are already happening, collecting encrypted data to unlock when quantum capabilities catch up. Cisco is turning tomorrow's threat into a plan enterprises can start building today: * New Quantum-safe communications advancements across Cisco's core portfolio. With a commitment to enable quantum-safe communications capabilities across the majority of Cisco's core portfolio by December 2026, Cisco is extending post-quantum protection to the systems where the most sensitive enterprise traffic flows. * Quantum-safe by default for new infrastructure. Starting today and moving forward, any newly introduced campus, branch and data center routers, switches, and firewall series will launch with quantum-safe secure boot. This builds on the same quantum-safe technology already shipping in campus smart switches. * The new Quantum Ready Assessments, available through Cisco IQ, identify the assets most exposed to "harvest now, decrypt later" attacks -- and where to start. Global availability planned for July 2026. * The new Quantum Resilience Framework gives enterprises a structured approach to post-quantum cryptography across two pillars: quantum-safe communications and quantum-safe products. Long-term resilience with Cisco To help customers navigate this new era, Cisco Services is announcing new capabilities:

Security Needs a New Operating Model

Security teams are not struggling because they lack tools. In most enterprises, the opposite is true. They have tools everywhere: firewalls, VPNs, identity systems, access controls, observability data, network telemetry, and cloud enforcement points. But together, they often create a new challenge: too many consoles, too much data, and too little shared context. The real opportunity is not adding another tool. It is reimagining how security work gets done: with shared context, governed action, and AI agents that help teams move at business speed. From fragmented tools to unified operations Security issues rarely stay inside one product boundary. A user blocked from an application might involve VPN posture, zero trust policy, branch connectivity, identity context, firewall rules, or application performance. An urgent policy change might require understanding years of firewall rules, business intent, and compliance requirements. A security gap might remain open because the signals needed to fix it are spread across disconnected systems. That fragmentation has a real business cost. Policy changes take longer than they should. Skilled administrators spend too much time decoding legacy rules or chasing knowledge that lives in someone's head. And when something breaks, teams lose time trying to answer a basic question: is this a security issue, a network issue, an access issue, or all of the above? Bringing security into Cisco Cloud Control That is the challenge Security in Cisco Cloud Control with AI Canvas is built to address. Security in Cloud Control brings Security Cloud Control into Cisco Cloud Control's unified operations platform. The goal is simple and powerful: give teams one governed environment to manage security enforcement points, correlate alerts across Cisco domains, and act with shared context. But the bigger shift is not just where the work happens. It is how the work happens. With Cisco Cloud Control, security becomes part of a broader operational model across domains. Inventory, topology, policy, identity, network, observability, collaboration, and data center context can come together in one shared experience. Security decisions are only as good as the context behind them. Here's what sets this model apart: * Unified context: Teams can connect security signals with network, identity, application, and infrastructure data. * Governed action: Operators can move from insight to action with validation and review built in. * Human-agent collaboration: AI agents can gather evidence, recommend next steps, and accelerate execution while humans stay in control. Making complexity workable with AI Canvas AI Canvas is where that complexity becomes workable. Inside AI Canvas, operators and AI agents investigate together in a shared workspace. They can correlate signals, build a timeline, trace dependencies, and move from question to resolution without losing context across handoffs. A Unified AI Assistant gives teams a natural language way to ask questions and get guidance. When an issue needs deeper investigation or execution, teams can escalate into AI Canvas with the right people, agents, and data in the same workspace. This matters because many of the hardest security problems are not purely technical. They are operational. Take firewall policy. A business team may need urgent access changed for an application, but the rules are poorly documented, the original admin has moved on, and nobody wants to make a risky change under pressure. In the old model, that means manual analysis, multiple reviews, and expert time spent reconstructing intent. In the new model, an admin can describe the desired policy in plain language. The system can translate that intent into candidate rules, map the change back to the business requirement, explain the reasoning, and validate the update before deployment. That does not remove the human from the decision. It gives the human better context, faster. Agentic operations for real security work The same pattern applies to zero trust. A team may not know that some users are reaching sensitive applications through VPN in a way that violates access best practices. Finding that manually could require continuous analysis across large volumes of event data. An AI agent can monitor for that pattern, surface the risk, explain why it matters, recommend configuration changes, and let an admin approve or refine the action. That is the practical value of agentic operations. Not AI for the sake of AI. Not automation without oversight. But AI that helps skilled teams move faster, preserve institutional knowledge, reduce risk, and spend less time stitching together fragmented evidence. Agentic operations rest on three core principles: * Human in the loop: Teams stay in control, bringing judgment and accountability to every important decision. * Cross-domain context: Agents connect signals across networking, identity, policy, observability, and enforcement points so teams see the full picture. * Purpose-built intelligence: Security workflows need agents that understand the domain, environment, and impact of every recommendation. This is how AI becomes more than an assistant. It becomes a collaborator. A new era for security operations Security in Cisco Cloud Control with AI Canvas is about bringing security work into the same operating model the rest of IT needs: unified context, governed action, and collaboration between people and AI agents. The end state is not a world where every task is automated blindly. It is a world where security teams can define once, enforce where needed, investigate with full context, and act with confidence. Security teams have never needed more tools. They need a better way to work. And with Security in Cisco Cloud Control with AI Canvas, that new operating model is here. We'd love to hear what you think! Ask a question and stay connected with Cisco Security on social media.