FortiAIGate on the Equinix Distributed AI Hub brings AI traffic inspection, policy enforcement and data protection closer to where enterprise models and workloads operate.
As enterprises move artificial intelligence from experimentation to production, securing AI across distributed infrastructure is becoming a critical requirement. Models, agents, applications, data and computing resources increasingly operate across public clouds, private data centres and edge environments, creating new challenges around visibility, governance, data protection and sovereignty.
Addressing this shift, Fortinet has announced the availability of FortiAIGate on the Equinix Distributed AI Hub through a validated reference architecture. The integration is designed to help organisations secure and govern AI interactions across hybrid and multicloud environments while maintaining control over sensitive data and where it is processed.
The architecture brings Fortinet’s AI security capabilities into Equinix’s distributed AI environment, which connects AI agents, models, data, compute resources and cloud providers through private, high-bandwidth, low-latency interconnection. FortiAIGate acts as an AI security guardrail, inspecting AI traffic, enforcing policies and providing visibility into interactions between users, applications, models and agents.
The move reflects a growing recognition that conventional perimeter security is not sufficient for enterprise AI. As workloads and data move across infrastructure boundaries, organisations need consistent controls over prompts, responses and model interactions, alongside the ability to identify unauthorised AI usage.
“Security needs to be built into distributed AI architectures from the start, enabling organisations to protect sensitive data, govern AI interactions and scale with confidence.”
— Anthony James, EVP of Marketing, Fortinet
“Enterprise AI is quickly moving beyond isolated applications to distributed environments where models, agents, data, and infrastructure span clouds and geographies. Security needs to be built into that architecture from the start,” said Anthony James, EVP of Marketing at Fortinet.
FortiAIGate is designed to inspect prompts and responses in both directions, apply centralised policies across sanctioned large language model and Model Context Protocol traffic, and provide controls over AI consumption. Organisations can also track per-token costs and establish budgets, limits and alerts as AI adoption expands.
Data sovereignty is another important consideration. The reference architecture is designed to keep control, management and inspection within customer infrastructure, supporting data residency and region-specific governance requirements. FortiGate also provides visibility into unsanctioned or shadow AI usage across the network, complementing governance of approved AI interactions through FortiAIGate.
Equinix says the distributed architecture gives enterprises flexibility to connect models, data and cloud resources while retaining control over their infrastructure and security choices. Maryam Zand, Vice President, Partnerships & Ecosystem Development at Equinix, said the validated architecture enables customers to apply FortiAIGate security and governance across connected AI environments.
Performance is also central to the proposition. Fortinet says the architecture can leverage GPUs, including those from NVIDIA, to support security inspection at the network layer without affecting AI workload performance.
The integration forms part of Fortinet’s broader AI-Native Security Fabric strategy, which brings together AI security, secure networking and infrastructure protection. Fortinet and Equinix also collaborate through FortiGate VM on Equinix Network Edge, extending their work into network security.
As enterprises scale agent-based applications and distribute AI workloads across locations, the ability to apply consistent security policies across infrastructure and AI interactions will become increasingly important. The Fortinet-Equinix architecture seeks to address that requirement by bringing governance closer to the environments where enterprise AI operates.
