Cisco (NASDAQ: CSCO) is extending its Cloud Control platform with AgenticOps capabilities intended to help partners and enterprise IT teams secure and manage AI-ready infrastructure from a shared operations layer. The update combines new access to network telemetry, automated background testing and AI-model services, giving teams a route to connect more operational context to troubleshooting and infrastructure-management workflows.
The development arrives as organizations place more AI workloads, agents and high-density wireless environments on networks that must remain available and govern access carefully. Cisco said Cloud Control gives teams and AI agents a common platform to connect, protect and run infrastructure across IT domains. The practical aim is to reduce the separate handoffs among network data, risk assessment and routine operations when an environment changes or a potential fault emerges.
A central component is the general availability of Cisco IQ for Partners. Cisco said the service gives participating partners access to customer-estate intelligence through an MCP server and expanded APIs, so they can integrate that data into their own managed services, workflows and AI agents. Customers can use role-based access control to grant a partner secure visibility into an installed base, while the partner uses current data to identify risks, plan upgrades or investigate an issue.
The company said Cisco IQ has visibility across nearly 20 million devices since April 2026 and more than 16,000 customers. Cisco is also adding quantum-safe and software end-of-vulnerability-support playbooks, priority-bug and software-conformance assessments, and tighter Cloud Control integration. The latter is designed to surface risk intelligence, including quantum-readiness status, and enable one-click action on eligible controller-managed devices. Those functions do not remove the need for change controls, but they can shorten the path from identifying a condition to acting on it.
Cisco is introducing Model Context Protocol servers for networking platforms as part of the AgenticOps work. MCP provides a standardized way for AI tools to access live network data. According to the company, the servers can reduce the custom connections partners would otherwise need to maintain and make it simpler to build troubleshooting and reporting workflows that join telemetry to experience metrics.
Cloud Control is also gaining Model Foundry, a platform service for discovering, evaluating, fine-tuning and running Cisco’s purpose-built models on an organization’s own data. Controlled availability begins with Cisco partners. For organizations using AI assistants in operations, that distinction matters: a model may be able to interpret infrastructure signals, but its usefulness depends on what authorized data and actions the surrounding platform exposes.
Other additions are aimed at continuous visibility. Cisco said it is integrating ThousandEyes into eligible Wi-Fi 6E and Wi-Fi 7 access points so the service can run automated background tests around the clock, including when users are offline. The company is also adding AI-assistant troubleshooting informed by device diagnostics and Technical Assistance Center knowledge, along with case-management functions inside the assistant.
The update is broader than a new dashboard. It puts partner operations, live telemetry and agent access on the same control plane, although enterprises will still need to define which actions agents may take and who approves changes. As AI increases the speed and volume of infrastructure activity, Cisco is positioning Cloud Control as an alternative to the older model of stitching together separate monitoring, support and remediation tools after an incident occurs.