Trend Micro announced on October 28th that it is integrating with NVIDIA BlueField. The goal is to make AI factories safer and more scalable by adding protection right at the data center layer. NVIDIA BlueField is a type of processor called a Data Processing Unit, designed to take over infrastructure and cybersecurity tasks so the CPU can focus on other work.
With this integration, Trend Vision One AI Factory EDR runs on NVIDIA BlueField to collect and monitor host and network information. It then checks that data against Trend Micro’s threat intelligence to spot suspicious activity. Trend Micro has also validated its solution on NVIDIA RTX PRO servers, offering enterprise-grade security for AI factories, including those run by federal agencies and other organizations that need high levels of trust.
Trend Micro is part of NVIDIA’s AI Factory for Government Reference Design, which gives full guidance for running AI workloads while staying compliant. Trend Vision One AI Application Security, or AI Guard, is now built into NVIDIA NeMo Guardrails, which is part of NVIDIA’s NeMo framework. This gives a scalable way to set up security rails that make sure AI interactions are safe, secure, and accurate.
Also Read: NRI Secure Launches ASMimosa to Boost Cyber Defense
For security teams protecting AI factories, this integration simplifies how they define, test, and manage AI guardrails. They can cover multiple types of AI tasks, map guardrails to organizational policies, and protect against risks like prompt injection, data leaks, misuse of tools or agents, jailbreaks, and hallucinations. The platform collects telemetry from guardrails so teams can see what’s happening, score risks, respond to incidents, and automatically update policies and playbooks across AI systems and the cloud.
This arrangement in general assists in identifying credential dumps, reverse shells, and other sophisticated threats. It enhances the protection of agent-based AI in content moderation, security, and privacy which are the three main areas.

