Fortinet has made FortiAIGate available through a reference architecture for Equinix Distributed AI Hub, giving enterprises a way to inspect AI traffic across dispersed infrastructure, the company said Nov. 5.
Fortinet says FortiAIGate checks prompts sent to AI systems and the responses they generate, then applies centralized rules to sanctioned large language model (LLM) and MCP traffic. Administrators can use it to monitor costs by token and set budgets, usage limits and alerts. The company says its rationale is that oversight, data protection, regional rules and performance are harder to manage when AI runs across different sites.
Fortinet says customers can keep the functions that oversee and check AI traffic on their own infrastructure to support data-residency requirements. It also claims AI workloads suffer no performance impact when security inspection uses GPUs, including NVIDIA hardware. The announcement gave no test conditions for that claim.
Equinix announced the Distributed AI Hub on March 11, 2026. The company said the hub was available at 280 data center locations and linked AI infrastructure providers through private connections designed for low latency. Equinix described it as vendor-neutral and named Palo Alto Networks as its first security integration at launch.
Fortinet said its existing network-security work with Equinix includes FortiGate VM on Equinix Network Edge. Maryam Zand, Equinix vice president of partnerships and ecosystem development, said in the announcement, “AI isn’t centralized.”
The Fortinet announcement named no customers or deployment figures and disclosed no financial impact. It gave no pricing or expected revenue and did not identify who would pay for the architecture or quantify commercial gains. The announcement named no scheduled follow-up milestone.
