Cisco’s new cloud platform aimed at securing AI infrastructure

Cisco Systems Inc. introduced Cisco Cloud Control, a cloud platform aimed at securing and managing AI infrastructure for enterprises. Announced at Cisco Live in Las Vegas, the platform unifies networking, security, observability, and collaboration tools into a single operational environment, supporting both human administrators and AI agents as digital coworkers. The platform addresses challenges created by agentic AI, which continuously interacts with enterprise systems and generates vast telemetry data. Cisco identified three major barriers to scaling agentic AI: infrastructure limitations, trust issues around autonomous systems, and data explosion from AI-driven operations. Executives noted risks like coordinated machine-speed attacks, citing concerns raised by Anthropic PBC’s Claude Mythos. Cloud Control serves as the foundation for Cisco’s AgenticOps model, shifting IT and security operations toward autonomous, agent-driven orchestration with human oversight. The platform consolidates Cisco’s networking, security, computing, and observability tools into a unified management plane, powered by a single sign-on access to Cisco Data Fabric, built on Splunk’s log analysis technology acquired in 2022. The platform integrates cross-domain telemetry with AI models tailored for networking and cybersecurity, alongside frontier AI models where applicable. AI agents within Cloud Control automate workflows, including problem identification, root-cause analysis, and recommended fixes, with testing on digital twins of customer environments and human validation. Cisco also launched Cloud Control Studio, enabling enterprises to create custom AI-powered workflows and agents. The platform ensures explainability, policy awareness, and compliance with guardrails, aiming to balance automation with human control. Executives emphasized the need for a collaborative model where agents handle operational tasks while humans retain oversight of critical decisions.