Cisco presented at Cisco Live a new unified platform to operate, protect and automate critical infrastructures in a context marked by agentic AI, with the aim of humans and artificial intelligence agents working on the same data layer and with permanent human supervision. The proposal, called Cisco Cloud Control, is proposed as the core of the company’s AgenticOps operating model and is accompanied by new security capabilities, quantum resilience and support services aimed at an increasingly faster and more automated threat scenario.
Cisco bets on a new operational layer
Cisco maintains that the speed of AI agents forces organizations to manage and defend their infrastructure “at the speed and scale of machines.” With that framework, Cloud Control centralizes networking, security, computing, observability and collaboration in a single view, with a single login and a common environment for human teams and autonomous agents.
The company presented the platform as a “command center” for agentic AI, from which critical operations can be created, monitored and coordinated without losing operational context. Among its objectives is to simplify the administration of complex environments and allow automatic decisions to always remain under human control.
Unified management and autonomous agents
One of the core elements of Cisco Cloud Control is cross-domain telemetry, which integrates data from networks, security, observability, and collaboration so that responses are made on a common basis. According to the company, this will help solve challenges such as service continuity, agent behavior and the efficient use of digital resources.
The platform also incorporates custom-designed models, including Deep Network Model, trained with 40 years of Cisco operational data. The company claims that this combination allows intelligence to be adapted to the specific problem and not only to the size of the model.
In addition, users will be able to work with trusted agents who follow a structured flow: detect incidents, identify causes, execute corrections, test changes before applying them and confirm the result. Cisco also highlighted features such as Expanded Experience Metrics, Deep Reasoning, Digital Twin, and Cisco Agentic Workflows to automate network operations without giving up visibility.
Spaces to create and customize
Cisco accompanied the platform with Cisco AI Canvas, a generative, multi-user workspace in which humans and agents share information in real time to investigate and resolve complex incidents. The idea is to preserve the context throughout the process, avoiding repetitions and loss of information between participants.
The other key piece is Cloud Control Studio, the design environment from which customers will be able to create custom agents and applications. Agent Builder will allow agents to be tailored to specific policies and workflows, connecting to more than 50 third-party platforms using native connectors or the open Model Context Protocol (MCP).
App Builder, for its part, will allow you to generate applications and workflows from natural language prompts, with coding assistance based on OpenAI Codex. Everything developed in Studio, together with the Cisco ecosystem, can be published in the Cloud Control Marketplace.
Security in real time
Cisco frames these announcements in a transformation of cybersecurity where reactive defense is no longer enough because the time between vulnerability and exploitation has been drastically reduced. The company says that as part of testing efforts with advanced AI models, it is subjecting its own products to early analysis to detect weaknesses before attackers do.
In that context, Cisco made the Foundry security specification open source so that other defenders can apply the same rigor to their AI-based security assessments. The company also assured that Cloud Control will function as a security command center to evolve defense in real time.
Among the new features is Live Protect, described as a digital immune system for Cisco products, capable of protecting against newly discovered vulnerabilities without reboots or maintenance windows. The feature is already available on Nexus 9000 switches and will be expanded to more products.
Protection for agents and hybrid networks
Cisco also deepened protection for environments where AI agents coexist with people and business systems. At the last RSA Conference, the company had already announced innovations to protect the world’s agents and the world of agents, and now adds improvements in areas such as AI Defense, Zero Trust for agents and Agentic SOC.
Along with this, Hybrid Mesh Firewall extends unified protection across networks, applications and own and third-party firewalls, with the aim of reducing the impact when an incident arises. The strategy aims at a more distributed perimeter, where security is applied consistently even if the infrastructure is hybrid and multi-provider.
Turn towards the quantum
Another of the axes of the announcement is resilience in the face of quantum computing. Cisco warned that “harvest now, decrypt later” attacks are already underway and that data encrypted today could be decrypted when quantum capabilities mature.
To respond to that threat, the company said it will expand secure communications capabilities against quantum computing across much of its core portfolio. Additionally, all new enterprise and data center routers, switches, and firewalls will include quantum-enabled secure boot capability.
Cisco also announced Quantum Ready, an assessment that identifies assets most exposed to such attacks, and a quantum resilience framework based on two pillars: secure communications and products secure against quantum computing. According to the company, the assessments will be available globally starting in July 2026.
Expanded services and support
The company rounded out the announcement with new features in Cisco Services and Cisco IQ. These include Resilient Infrastructure, a three-step service to assess exposure, modernize infrastructure and strengthen defense against advanced threats.
Cisco IQ will also incorporate a resilient infrastructure playbook, support for on-premises deployments aimed at data sovereignty, and the Peer Benchmarking tool, which uses anonymized data to compare risk exposure and support status with similar organizations.
With this set of launches, Cisco seeks to position itself as a reference provider in the transition towards the agentic company, a model in which automation stops being peripheral and becomes part of the operational and security center of the organization.
