Graphics and technology developer Nvidia has grown considerably since its founding in 1993 and is now the dominant leader in Graphics Processing Units (GPUs) for artificial intelligence, with 92% market share. The company also designs high-performance chips and software platforms, powering AI training and inference for data centers, autonomous vehicles, professional visualization, and gaming.
This week, the Santa Clara, California-based firm announced it is expanding its technological footprint into cybersecurity, specifically AI-powered cybersecurity. It will extend Zero Trust principles into Operational Technology (OT), embedding AI-driven protection directly into key infrastructure, including industrial control systems (ICS) across energy, manufacturing, transportation, and utilities.
Those are areas increasingly dependent on enterprise networks and the cloud, but also increasingly exposed to cyber threats.
“Unlike traditional IT environments that manage data and applications, OT systems control real-world processes where cyber incidents can have immediate consequences for safety, availability and operational continuity,” Nvidia explained in a blog post.
It added that many of the OT systems were also designed with reliability and longevity in mind, not to address the latest threats.
“This can widen the gap between modern attacks and existing defenses. Even as OT and ICS environments modernize with improved automation, connectivity and analytics, most were not built to withstand adaptive, software-driven cyberattacks that evolve in real time,” the GPU developer explained.
Collaboration With Cybersecurity Providers
To address emerging threats, Nvidia announced it would collaborate with leading cybersecurity providers, including Akamai, Forescout, Palo Alto Networks, and Xage Security. It will further work with industrial automation innovator Siemens.
The goal is to deliver accelerated computing and AI to OT cybersecurity, advancing real-time threat detection and response across critical infrastructure.
Nvidia suggested that it would “represent a fundamental shift in OT and ICS cybersecurity” by embedding security into and distributing it across critical infrastructure. The company’s BlueField Data Processing Units (DPUs) are at the heart of this initiative, as the specialized processors can run the security services on dedicated hardware, thus keeping protection separate from operational systems “so critical processes remain unaffected.”
What the Partners Are Providing
The “dream team” of companies involved is each contributing to the effort, with Siemens and Palo Alto Networks embedding security into industrial automation. Later this year, Siemens will demonstrate its AI-ready Industrial Automation DataCenter to showcase how it can deliver an AI-ready, zero-trust solution tailored to industrial automation.
The Prisma AIRS AI Runtime Security from Palo Alto Networks will further provide continuous monitoring of industrial traffic, with Akamai Technologies’ Guardicore Platform enabling agentless segmentation to isolate applications, and devices or workloads into controlled security zones.
Zage Security will help Nvidia deliver zero-trust security to both energy infrastructure and any AI systems it supports.
“Across these environments, a consistent OT cybersecurity architecture is taking shape,” the blog post continued. “Security services run at the edge on Nvidia BlueField DPUs, close to the operational systems they protect.”
Addressing a Compute Problem
Ensar Seker, CISO at SOCRadar, told ClearanceJobs that Nvidia’s announcement is absolutely a step in the right direction.
“Critical infrastructure security is increasingly an AI and compute problem, not just a tooling problem,” said Seker. “We’re dealing with massive telemetry volumes from OT, IT, cloud, identity systems, and supply chains. Traditional SIEM (Security Information and Event Management) and rule-based detection simply cannot keep pace with the scale or speed of modern threats.”
Nvidia’s positioning of AI infrastructure directly into cybersecurity workflows further signals recognition that defending power grids, healthcare systems, transportation networks, and financial platforms requires high-performance AI pipelines, not incremental patchwork.
The GPU developer is also well-positioned to deliver acceleration, both technically and strategically.
“First, at the infrastructure layer, they provide the GPU compute backbone required to train and run large-scale detection models, behavioral analytics engines, and AI agents for SOC automation,” said Seker. “Second, their AI frameworks – like inference optimization, digital twin simulation, and accelerated data processing – can enable real-time anomaly detection in industrial environments where milliseconds matter. Third, Nvidia sits at a unique intersection of AI, edge computing, and data center architecture, meaning they can help move advanced detection closer to where critical systems actually operate.”
However, the real impact will still depend on execution.
This effort needs to be more than just another AI-powered marking label; otherwise, it won’t materially change outcomes.
“If Nvidia helps security vendors operationalize large-scale behavioral modeling, OT-aware AI agents, and high-fidelity threat simulation environments, then this could meaningfully raise the defensive baseline for critical infrastructure. The key will be integration, AI must plug directly into SOC workflows, not operate as an isolated innovation layer,” Seker continued.
He further told ClearanceJobs that strategically, this also reflects something bigger.
“Cybersecurity is no longer just a software domain, it is now compute-driven,” Seker added. “The same AI acceleration that powers autonomous vehicles and generative AI models can be redirected toward defending nation-state-level targets. That convergence is where the real opportunity lies.”
