{"pageModel":{"attributes":{"id":"","name":"97384.dita","viewName":"DitaDetail"},"elements":{"ditaContent":{"name":"DITAContent","value":"
\r\n

AI Defense Overview

\r\n
\r\n
\r\n

Manage AI risks in a multi-cloud, multi-model world

\r\n

AI Defense addresses the risks introduced by the development, deployment, and use of AI. It combines AI-specific detection and defense measures with the existing network visibility and enforcement points in the Cisco Security Cloud.

\r\n

AI Defense addresses these key areas in AI security:

\r\n
    \r\n
  • \r\n

    AI cloud discovery and visibility: Discover AI models and AI assets running across your AWS, Azure, and GCP clouds.

    \r\n
    • \r\n
  • \r\n

    AI model and application validation: Cisco AI Defense validation provides automated red-teaming that detects safety and security vulnerabilities in your AI models and AI applications, enabling you to configure AI Defense runtime guardrails.

    \r\n
    • \r\n
  • \r\n

    AI runtime protection: Cisco AI Defense Runtime protection guards AI applications and their users by inspecting inbound and outbound AI application traffic and blocking or alerting on prompt injection (PI) attempts, personally identifiable information data leakage (PII leaks), safety risks, and security risks.

    \r\n
    • \r\n
  • \r\n

    AI supply chain security: Cisco AI Defense Supply Chain Security scans model files, repositories, MCP servers, and agent traffic to identify malicious components and mitigate and address vulnerabilities.

    \r\n
    • \r\n
  • \r\n

    AI access protection: The Cisco Secure Access-AI Defense integration protects employees using 3rd-party AI apps with data loss prevention (DLP) and guardrails.

    \r\n
    • \r\n
  • \r\n

    Network fabric-level AI guardrails: The Cisco Multicloud Defense-AI Defense integration lets you enforce AI safety and security policies at the network fabric level for cross-cloud traffic.

    \r\n
    • \r\n
  • \r\n

    AI Defense hybrid deployment for your private environment: Scan and protect AI models and applications locally in your private cloud while keeping the ease of management of the standard AI Defense UI that runs as a cloud service.

    \r\n
    • \r\n
\r\n
\r\n
\r\n

Part of Cisco Security Cloud Control

\r\n

AI Defense is part of the suite of products that you manage from Cisco Security Cloud Control. Security Cloud Control is the single integrated interface for all Cisco security products. Using the Security Cloud Control navigation in the platform menu and main menu bar of the portal, you can access products including AI Defense, Multicloud Defense, and Secure Access.

\r\n

Security Cloud Control provides common facilities for managing users, credentials, integrations, and product subscriptions. For example, Security Cloud Control is where you can generate the API keys used for the AI Defense APIs and other Cisco security products. See the Overview of Cisco Security Cloud Control for more details.

\r\n
\r\n
\r\n
\r\n","ditaVal":"","format":"html"},"bookTitle":{"value":""},"shortDescription":{"value":""}}},"parameters":{"appId":"SccAiDefense","topicAlias":"overview"}}