Fortinet announced a deepened integration with Nvidia, touting AI security and strengthened sovereignty for users.

The enhanced FortiAIGate platform combines Fortinet’s security technologies with Nvidia’s accelerated computing stack to provide real-time monitoring, governance, and protection for AI workloads. The offering is designed to secure AI deployments without introducing significant latency or disrupting inference performance.

Fortinet claimed the platform can inspect and govern AI traffic in-line between applications and AI models, enabling organizations to monitor prompts, responses, and interactions involving AI agents and model context protocol (MCP) servers. This is achieved through runtime security controls and guardrails for large language models (LLMs), including protections against prompt injection attacks, toxic content generation, and unauthorized data exposure. The platform also logs suspicious prompts and responses to assist with compliance and forensic analysis.

The company also highlighted AI sovereignty capabilities. As explored in SDxCentral’s Sovereignty Supplement, governments and enterprises are increasingly scrutinizing where AI training and inference data reside. This has seen the likes of NTT tout sovereignty at the on-premises, graphics processing unit (GPU) layer through the aid of all-photonics network solutions that offer similar latency to the cloud without the risk of data leakage.

With FortiAIGate, organizations can deploy the solution using self-hosted infrastructure to ensure AI workloads and sensitive data remain within national borders and comply with local regulations such as the European Union’s General Data Protection Regulation (GDPR).

The platform can be deployed across on-premises, cloud, hybrid, and edge environments as either a GPU-powered appliance, virtual appliance, or containerized deployment on Nvidia-certified systems.

Performance optimization is another focus of the partnership, with the platform leveraging Nvidia's Blackwell GPUs and the Nvidia Dynamo distributed inference-serving framework to accelerate AI security processing while reducing reliance on central processing unit (CPU)-based infrastructure. The company claims this approach lowers power consumption and total cost of ownership compared with traditional security architectures.

FortiAIGate also supports multitenant AI deployments using Nvidia virtualization technologies such as multi-instance GPU (MIG), allowing multiple isolated AI workloads to run on shared GPU infrastructure with guaranteed quality of service and fault isolation.

“Enterprises everywhere are racing to adopt AI, and security has become a critical enabler of that innovation,” Fortinet COO John Whittle noted. “Together with Nvidia, we’re delivering a solution that helps organizations secure and optimize AI deployments while maintaining performance, controlling costs, and meeting data sovereignty requirements.”

“The accelerating shift toward autonomous AI agents is creating unprecedented demand for secure, high-performance enterprise computing platforms,” Justin Boitano, VP for enterprise AI platforms at Nvidia, added. “By integrating its FortiAIGate solution with the full-stack Nvidia AI platform, Fortinet provides zero-trust security and real-time governance, reducing threat exposure by shortening response times.”

The solution follows a December tie up between the pair, where Fortinet unveiled an integrated security solution developed with Nvidia, embedding firewall and zero-trust capabilities directly into accelerated AI infrastructure.

That integration enabled Fortinet’s FortiGate VM virtual firewall to run natively on Nvidia’s BlueField-3 data processing unit, offloading core security functions, such as firewalling, network segmentation, and policy enforcement. It's designed to protect high-performance data centers without slowing down work.