Core Insights - Cisco is introducing significant security innovations aimed at the agentic AI ecosystem, focusing on enabling organizations to deploy AI agents confidently at scale [1][2][3] Security Innovations - Cisco's new solutions address AI security issues by establishing trusted identities, enforcing Zero Trust Access controls, and providing tools for security operations teams to counter threats rapidly [1][5] - The company is extending Zero Trust Access to AI agents, integrating new capabilities in Identity and Access Management (IAM) and policy enforcement to enhance visibility and governance over agentic actions [5][6] AI Agent Adoption - A Cisco survey revealed that 85% of major enterprise customers are experimenting with AI agents, but only 5% have implemented agentic technology in production [2] Key Security Pillars - Cisco is focusing on three main pillars for securing the agentic workforce: protecting the world from agents, protecting agents from external threats, and enabling rapid detection and response to AI incidents [3][6] - The company emphasizes the need for onboarding AI agents to establish their identity and accountability, which is currently lacking in many enterprises [3] AI Defense Tools - Cisco is launching AI Defense: Explorer Edition, a self-service solution that allows organizations to test the resilience of AI models against attacks before deployment [8][12] - The introduction of DefenseClaw, an open-source secure agent framework, aims to automate security processes and integrate with NVIDIA's OpenShell for enhanced security during agent deployment [13][14] SOC Enhancements - Cisco is evolving the Security Operations Center (SOC) from reactive to proactive by embedding AI capabilities into workflows, allowing analysts to prioritize threats more effectively [15][17] - New specialized AI agents are being introduced to automate security workflows, enabling the SOC to operate at machine speed and scale [17]