Cisco, a US-based enterprise networking, security and infrastructure company, has introduced Cisco Cloud Control, a platform designed to let human operators and AI agents manage, monitor and defend IT infrastructure within the same environment. The company unveiled the platform at Cisco Live and described it as the foundation for what it calls its AgenticOps operating model.
Cisco Cloud Control brings together a customer's networking, security, compute, observability and collaboration data under a single login and a single view. According to Cisco, people and AI agents work from the same data layer and the same operational context, with customers able to build their own applications and agents inside the platform using natural language.
The platform connects to a broader set of third-party tools, including Amazon Web Services, the project management platform Linear, Microsoft, the incident response platform PagerDuty, ServiceNow, Slack, and Google Cloud, which Cisco said now includes Wiz, a cloud security platform.
Jeetu Patel, President and Chief Product Officer at Cisco, described the shift the platform is meant to address:
"AI agents reason and act continuously at software speed, and that changes everything about how we scale, manage, and defend our critical infrastructure. 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."
The platform is built around several components. Cross-domain telemetry pulls data from networking, security, observability and collaboration systems into Cloud Control so that humans and agents can act on shared information covering uptime, agent behavior and token usage. For reasoning, the platform draws on a mix of purpose-built and frontier AI models, including Cisco's Deep Network Model, which the company built on 40 years of its own operational networking data.
Cisco said the platform also includes agents capable of following a defined sequence from detection to resolution, identifying problems, determining causes, implementing fixes, testing changes before deployment and confirming that user experience has recovered. These agents draw on Cisco's telemetry and purpose-built models, along with additional capabilities the company calls Expanded Experience Metrics, Deep Reasoning, Digital Twin and Cisco Agentic Workflows. Teams can automate network operations through what Cisco describes as an agentic loop, while keeping the resulting actions visible and subject to governance.
Cisco AI Canvas is a shared workspace where operators and agents investigate and resolve issues together using the same live evidence in real time, with context persisting across shifts so information is not lost or repeated between sessions. Cloud Control Studio gives customers two ways to customize the platform. Agent Builder lets customers build their own agents tailored to internal policies and workflows, connecting to more than 50 third-party platforms through native connectors or the open Model Context Protocol. App Builder lets customers build and publish applications from natural-language prompts, using OpenAI Codex, an agentic coding platform from OpenAI, the AI research company that also develops ChatGPT. Everything built within Studio, along with agents and apps from across Cisco's ecosystem, can be published to the Cloud Control Marketplace.
Cisco Cloud Control is entering Controlled Availability in the United States starting today, with global availability to follow.
Cisco is pairing the platform's launch with a set of security announcements tied to what it describes as a shrinking gap between the discovery of a vulnerability and its exploitation, a window the company says AI has compressed from weeks to minutes. Cisco is a charter member of Project Glasswing, an initiative from Anthropic, the AI safety and research company that develops the Claude family of models, and of Daybreak, a program from OpenAI. Through these programs, Cisco said it stress-tests its own products using frontier AI models to find weaknesses before attackers do. The company has also open-sourced its Foundry Security Spec, making the same evaluation approach available to other defenders.
Among the new protections, Live Protect is designed to shield supported Cisco products from newly discovered vulnerabilities at runtime without interrupting operations through reboots or upgrades. The capability is now available in Cisco's N9000 series switches and included with the Nexus One product entitlement. Cisco plans to extend Live Protect to additional products over the coming months, starting with campus and branch smart switches and followed later in the year by secure routers.
Cisco also introduced Hybrid Mesh Firewall, which extends unified protection across networks, applications, and both Cisco and third-party firewalls to limit the impact when a breach occurs. Separately, Cisco said it is expanding the agentic security capabilities it first announced at the RSA Conference, with updates spanning AI Defense, Zero Trust protections for AI agents, and what the company calls its Agentic SOC.
On quantum security, Cisco said it is responding to a category of attacks known as harvest now, decrypt later, in which adversaries collect encrypted data today with the intention of decrypting it once quantum computing capabilities mature. The company has committed to enabling quantum-safe communications across most of its core product portfolio by December 2026.
Starting today, any newly introduced campus, branch and data center routers, switches and firewalls will ship with quantum-safe secure boot by default, building on quantum-safe technology already present in its campus smart switches. Cisco also introduced Quantum Ready Assessments, delivered through Cisco IQ, to help customers identify which assets are most exposed to harvest now, decrypt later attacks, with global availability planned for July 2026. A separate Quantum Resilience Framework gives customers a structured approach to post-quantum cryptography across two areas, quantum-safe communications and quantum-safe products.
Cisco is also expanding its services business to address what it calls frontier model risk. Resilient Infrastructure Services, delivered through Cisco Support and Professional Services, follows a three-step approach covering exposure assessment, infrastructure modernization and defense resiliency. Cisco IQ, now fully integrated into Cisco Cloud Control, serves as Cisco's AI-powered delivery system for its support and professional services, built around a Resilient Infrastructure Playbook that combines AI-driven insights with Zero Trust principles. Cisco IQ will now support on-premises deployment for customers with data sovereignty requirements, and its Peer Benchmarking feature uses anonymized data to compare a customer's Last Day of Support risk exposure and security vulnerability rates against organizations of similar size, sector or infrastructure profile.
Global availability for these capabilities is planned for July 2026.
