Author: Vivek Doshi, Director of Product Management, Xage Security
We had an incredible time at 2025 NVIDIA GTC in Washington, D.C., exploring what’s next on the AI frontier — from secure AI factories and accelerated computing to robotics and autonomous mobility. One key takeaway was NVIDIA’s growing focus on AI for critical infrastructure, from national energy systems to industrial operations, and the urgent need to secure these environments as they become faster, more autonomous, and more interconnected.
NVIDIA and its technology partners are pioneering scalable AI infrastructure that integrates GPUs, BlueField DPUs, and advanced AI software into full-stack systems capable of processing massive volumes of data at unprecedented speeds. These AI “factories” are rapidly becoming the new critical infrastructure, powering breakthroughs in energy, transportation, manufacturing, and defense.
But as AI becomes central to national and industrial systems, security must evolve in lockstep. The question isn’t just how far AI can go, but how we ensure it doesn’t fall into the wrong hands. In this new era of intelligent systems, defense-in-depth is essential, yet security must operate inline so it enables performance, not hinder it.
Xage and NVIDIA: Zero Trust for the AI Age
At the conference, Xage Security announced a strategic partnership with NVIDIA, including integration with the BlueField DPU. Together, we’re delivering a Zero Trust solution purpose-built to protect AI factories and critical infrastructure without compromising speed or performance.
Xage enforces least-privilege access controls across every stage of the AI lifecycle from training and inference to agent operations. By isolating AI workloads and applying line-rate segmentation, Xage prevents lateral threat movement and delivers full visibility into packet flows.
With identity-based Zero Trust enforcement running at line speed on BlueField, organizations can:
- Limit the actions users and AI agents are authorized to take
- Prevent unauthorized privilege escalation and data leakage
- Ensure AI agents remain trustworthy and compliant as they scale and evolve
By leveraging BlueField’s hardware acceleration, Xage extends identity-based access control and segmentation across massive data centers and mission-critical environments, ensuring exceptional performance, scalability, and resilience.
With Zero Trust enforcement across humans, systems, AI models, and agents, Xage provides a complete, closed-loop security architecture designed for the next generation of critical systems. The future of AI is fast, distributed, and intelligent, and it must be secure by design.
Learn more about the Xage and NVIDIA integration →
Critical Infrastructure and AI Converge
NVIDIA’s commitment to advancing AI for critical missions was on full display at GTC. During the conference, NVIDIA and Oracle announced plans to build the U.S. Department of Energy’s largest AI supercomputer, designed to accelerate scientific discovery. This initiative underscores how AI is becoming fundamental to national infrastructure and energy innovation.
But with that progress comes new responsibility: securing the AI systems that will power the world’s most essential services. These environments, from DOE research facilities to industrial control networks, require real-time, Zero Trust security to ensure resilience and integrity.
At Xage, we see the same momentum across our customer base. Critical infrastructure operators are adopting AI at the edge and in the cloud to drive efficiency, reliability, and innovation. As NVIDIA builds the high-performance backbone for AI innovation, Xage ensures that backbone is protected end-to-end.
AI innovation and critical infrastructure are converging, and they must be secured together. With NVIDIA and Xage working side by side, organizations can accelerate safely into the AI-driven future, where innovation and security operate at the same speed.
