Autonomous AI agents capable of planning and executing tasks independently face a critical security gap that traditional models can't address. TrendAI announced expanded collaboration with NVIDIA this week to secure the chipmaker's new OpenShell runtime for agentic AI, bringing enterprise-grade security directly into autonomous agent architectures.
The partnership targets NVIDIA OpenShell, an open source runtime introduced at GTC 2026 for long-lived, self-evolving agents with planning, memory, and tool execution capabilities. While these systems promise significant productivity gains, they introduce risks including unauthorized skills access, hidden behaviors, prompt injection attacks, and unintended system access that conventional security approaches weren't designed to handle.
"Agentic AI changes the security equation," said Rachel Jin, Chief Platform and Business Officer at TrendAI. "When AI systems can plan, take action, and interact with other tools on their own, the risk profile looks very different from traditional AI."
Traditional AI security models were built for short-lived interactions between users and models. Agentic systems operate continuously across environments, creating new attack surfaces that require runtime enforcement rather than perimeter-based protection.
TrendAI Vision One's layered security architecture for OpenShell provides centralized governance enforced directly in the agent runtime along with continuous scanning of agent skills and MCP integrations. Dynamic behavioral analysis detects hidden or malicious actions while inline policy enforcement blocks untrusted skills at runtime.
The solution includes AI-specific threat protection against prompt injection and sensitive data exposure detection plus continuous monitoring through agentic telemetry and SIEM integration. Organizations gain the ability to define trust boundaries while maintaining visibility across autonomous agents without limiting innovation.
NVIDIA's commitment to agentic AI extends beyond hardware infrastructure. The company is investing $26 million over five years into its open model initiative according to Wired reports confirmed by NVIDIA executives earlier this week at GTC 2026.
"Agentic AI opens the door for a new class of applications that can plan, reason, and take action," said Pat Lee, vice president, Strategic Enterprise Partnerships at NVIDIA. "By working with TrendAI™, we're helping developers add visibility and controls to make it safer to run autonomous AI."
The collaboration extends across NVIDIA's AI-Q blueprint and NeMo Agent Toolkit enabling consistent security as agentic systems scale across enterprise environments including AWS Google and Microsoft cloud platforms.
TrendAI operates globally with 6,000 experts across 75 countries protecting organizations from hundreds of millions of threats daily through its unified cybersecurity platform.
