TrendAI has enhanced its collaboration with Nvidia to bolster security for agentic AI through Nvidia OpenShell, an open-source runtime for autonomous AI agents. This integration fosters security within the runtime architecture, essential for agents that operate autonomously across various environments. TrendAI Vision One will deliver a layered security architecture, incorporating governance and monitoring controls to manage agent behavior changes over time. As agentic AI gains traction among enterprises looking for automated workflows, it also introduces new security challenges distinct from traditional AI. Chief Platform Officer Rachel Jin emphasizes the need for robust security measures as these systems can independently execute tasks and interact with tools. The collaboration aims to improve visibility and control for developers, encompassing runtime security measures and AI-specific threat protection like prompt injection detection. This initiative aims to streamline enterprise adoption of agentic AI while maintaining compliance and effective risk management throughout the agent lifecycle.
Source link
