F5 & Equinix join forces on enterprise AI security

F5 has teamed up with Equinix to offer a combined system for governing and securing enterprise AI across hybrid and multicloud environments. The arrangement links F5 AI Guardrails with the Equinix Distributed AI Hub.

The joint offering is aimed at businesses running AI workloads across multiple models, clouds, agents and data sources, where management and oversight have become more difficult. It is designed to create a single control layer for AI interactions, running over private interconnects and applying common policies across different environments.

Large organisations are increasingly building generative and agent-based AI services across a mix of on-premise systems, public cloud platforms and external model providers. That spread has created new risks around data leakage, policy breaches, harmful outputs and limited visibility into how AI tools are being used.

F5's role centres on its AI Guardrails product, which applies policy-based controls during AI interactions. Equinix contributes its Distributed AI Hub, described as a unified framework for connecting AI infrastructure, cloud services, data platforms, security services and related suppliers through its data centre footprint.

Together, the system is intended to help customers avoid rebuilding existing architectures while still applying common rules to AI traffic. It can also support organisations with stricter data privacy and sovereignty requirements because F5 AI Guardrails can be deployed on-premise within Equinix facilities.

Managing sprawl

The announcement comes as many companies face rising costs and technical complexity from AI projects that have expanded beyond early experiments. Businesses often connect multiple models to internal and external data sources, while different teams deploy tools in parallel, creating fragmented systems and inconsistent controls.

That pattern has also contributed to so-called shadow AI, where employees or business units adopt AI services outside approved governance processes. In regulated sectors, the lack of central oversight can raise compliance concerns, especially when data crosses borders or is shared with third-party model providers.

Traditional application security tools have not always addressed those issues because they were built for conventional software and web traffic rather than AI prompts, model responses and agent-to-agent exchanges. The combined framework is intended to improve visibility and create records that support internal oversight and regulatory reporting.

Equinix said its platform gives enterprises access to a broad ecosystem through more than 280 interconnected data centres and over 10,000 customers. That structure allows users to connect with model companies, GPU cloud providers, network suppliers, data platforms and security services without locking themselves into a single provider.

For F5, the agreement extends its push into AI-specific security controls as demand grows for tools that monitor prompts, outputs and model behaviour. Its software can detect and block sensitive data from leaving controlled environments, identify policy violations and moderate outputs that do not meet an organisation's rules.

Executive comments

John Maddison of F5 said the joint work is intended to reduce friction for companies trying to move AI systems into production.

"Enterprises want to move fast with AI, but security gaps, fragmented governance, and compliance uncertainty keep slowing them down," said John Maddison, Chief Marketing Officer, F5. "Together with Equinix, we are giving organisations a way to deploy distributed AI across hybrid and multicloud environments with the speed and flexibility they need, while ensuring every AI interaction is protected, governed, and audit-ready. This is about making secure AI adoption the path of least resistance, not the bottleneck."

Brian Stein of Equinix said the arrangement reflects customer demand for neutral infrastructure that can connect multiple AI and cloud providers.

"Enterprises require infrastructure that supports AI innovation without forcing tradeoffs between performance, security, or control," said Brian Stein, Senior Vice President of Infrastructure, Equinix. "Equinix is the neutral platform where AI, cloud, and networking infrastructure converge. By combining F5 AI Guardrails with the Equinix Distributed AI Hub, we offer customers a simpler and smarter vendor-neutral foundation to build and scale distributed AI with confidence."

Use cases

The framework is intended for a range of enterprise uses, including preventing sensitive data from being exposed to large language models, applying consistent governance rules across dispersed AI deployments and limiting harmful or non-compliant outputs. It is also positioned as a way to give companies better control over AI spending and to align where AI processing takes place with local data residency requirements.

F5 added that its AI Red Team product can be used alongside the combined system to test for weaknesses in guardrail policies. The broader message from both companies is that AI deployment is becoming as much an infrastructure and governance challenge as a software one, particularly for businesses managing multiple clouds, multiple models and stricter regulation.

The combined system is centred on a single policy framework intended to govern AI interactions across the clouds, models and agents an enterprise uses.