Welcome to the first email in our 2026 Security Cloud Control Updates Series! Throughout this series, we’ll spotlight each product integrated with Security Cloud Control and the latest features deployed this year, including:

• AI Defense
• Multi-Organizations
• Security Cloud Control Firewall Management
• Security Cloud Control Platform
• Mesh Policy Engine

In this email, we start with AI Defense, Cisco’s solution for discovery, validation, and real-time security of AI models, applications, and users.

AI Defense 2026.1.1 Release – January 6, 2026

The AI Defense 2026.1.1 release introduces several key enhancements to help organizations secure AI systems more efficiently:

• AI Defense Adopts New AI Threat Taxonomy
• AI Defense MCP Server Catalog (Beta)
• AI Defense Hybrid Deployment on Azure

AI Defense Adopts New AI Threat Taxonomy

Cisco introduces the Integrated AI Security & Safety Framework, a comprehensive taxonomy to describe, manage, and mitigate AI-specific threats.

AI Defense now categorizes results and events according to this framework, giving organizations the ability to quickly understand and act on AI risks across:

• Runtime protection
• Validation
• Model scanning
• MCP scanning

The taxonomy organizes threats into four layers:

1. Objectives – the “why” behind attacks
2. Techniques – the “how”
3. Subtechniques – specific variations of techniques
4. Procedures – real-world implementation

For example:

• Attacker objectives: 19 types including goal hijacking, jailbreaks, communication compromise, data privacy violations, privilege escalation, harmful content generation, and cyber-physical manipulation.

• Techniques & subtechniques: 150+ tactics covering prompt injections, jailbreaks, multi-agent manipulation, memory corruption, supply chain tampering, environment-aware evasion, and tool exploitation.

By mapping threats this way, operational teams gain specificity needed to mitigate risks effectively.

AI Defense MCP Server Catalog

The MCP Server Catalog, now available as a beta feature, enhances AI Defense gateway scanning. Key capabilities include:

• Registration Endpoint: Easily register MCP servers with server URL, credentials, and type
• Scanning Endpoint: Launch on-demand or scheduled vulnerability scans

AI Defense MCP scanning analyzes tool descriptions to detect tool poisoning attempts and generates unique signatures to validate tool integrity. Any malicious or misleading content is blocked automatically.

AI Defense Hybrid Deployment on Azure

AI Defense now supports hybrid deployments on Azure, allowing AI Validation and Runtime processing to run in your Azure cloud environment.

• This reduces the need to publicly expose model endpoints
• The AI Defense control plane remains SaaS via Cisco Security Cloud Control

This deployment option is available in controlled availability, providing flexible, secure cloud options for organizations managing sensitive AI workloads.

Next in the Series:
In our next email, we’ll cover Multi-Organizations, highlighting the new features in Security Cloud Control that simplify management for MSPs and enterprises across multiple organizations.

Stay tuned!