Particle.news
Download on the App Store

Cisco Pushes Agentic AI Safety With AI Defense Update and Cloud Control Launch

The company is centralizing agent orchestration and safety so enterprises can govern automated agents across networks and cloud environments.

Overview

  • Cisco updated AI Defense with adaptive red teaming and a natural-language Policy Studio that lets teams simulate custom attacks and build context-aware guardrails for individual agents.
  • AI Defense now auto-discovers agents' full dependency graphs, catalogs assets in a central AI inventory, and allows developers to run vulnerability scans from CI/CD pipelines.
  • Cisco placed Cisco Cloud Control in controlled availability as a single AgenticOps control plane using AI Canvas for orchestration and telemetry, with a planned Cloud Control Studio for low-code agent building later this year.
  • The company previewed Network Actions, including an alpha digital twin and a next-generation switch/router platform to support network-specific agent controls, with a beta planned soon.
  • Cisco says the tools are platform-agnostic and integrate with major clouds, agent frameworks, NVIDIA, and 50+ partners, but analysts warn fierce vendor competition and the need to build multi-vendor trust before broad enterprise adoption.