Charting the Path to Governed Agentic AI with Nutanix Enterprise AI 2.8

Enterprise AI is quickly entering a new phase. Organizations are building and deploying AI agents to drive greater operational efficiency and automate multi-step workflows. Beyond answering questions, these agents are connecting to enterprise applications, accessing business tools, and retrieving information from external and internal sources. The impact compounds as a single autonomous agent can spawn multiple agents to complete tasks, each connecting to systems and business tools. 

This level of orchestration introduces critical governance gaps for organizations, with leaders asking key questions:

• Which systems should agents have access to?
• How do we govern agent interactions with our business tools consistently?
• How do we help ensure agents adhere to internal security standards?
• How do we maintain visibility at scale as agents operate within our business?

Gartner predicts that “by 2027, 40% of enterprises will demote or decommission autonomous AI agents due to governance gaps identified only after production incidents occur.”¹

The next phase of agentic AI will be measured by the ability to operate agents with governance, visibility, and control. That’s what we set out to solve with the latest Nutanix Enterprise AI (NAI) 2.8 release. This release continues to evolve NAI with new control, governance, and efficiency capabilities across Nutanix Agent Gateway and Nutanix Private Inference.

Governing the Agentic Ecosystem with Nutanix Agent Gateway

AI governance has evolved beyond simply managing individual models. Today’s enterprises require consistent control over agents, tools, models, and users. The adoption of the Model Context Protocol (MCP) allowed a standardized way for agents to connect to enterprise tools, removing the complexity of building custom integrations for each connection. As organizations deploy MCP servers in their environment, governance gaps can introduce security risks and create operational bottlenecks.

Today, we’re excited to announce the general availability of the MCP server management capability in Nutanix Agent Gateway, centralizing control for agent access and allowing agents to securely connect to tools and private data sources with strict governance. Nutanix Agent Gateway acts as a single, secure front door between AI agents and the tools they access, whether MCP servers are locally deployed in an NAI environment or remotely deployed. This capability delivers centralized governance and unified observability across agent-tool interactions, enabling:

• Secure tool deployment: customers can deploy local MCP servers directly inside their NAI environment, allowing their private tools and sensitive data to stay within their organization’s infrastructure perimeter and accessible to their agents. 
• Simplified integration at scale: a single unified endpoint aggregates multiple MCP servers, eliminating the need for developers to build custom connectors for every tool.
• Fine-grained control: take total control over tool permissions by assigning capabilities, such as read-only vs. write, for specific users or API keys.
• Zero-downtime operations: locally deployed MCP servers now support rolling updates, keeping your tools fully accessible while upgrading behind the scenes.

MCP Server Management in Nutanix Agent Gateway

In addition to MCP server management, NAI 2.8 also introduces enhancements to manage token budgets:  

Header-Based Rate-Limiting with Token Budgets (Tech Preview): enforce granular control on token budgets for individual users without changing existing user management​ through Nutanix Agent Gateway.

Scaling Nutanix Private Inference with Greater Efficiency 

As AI adoption grows and workloads become more demanding, enterprises need infrastructure that can scale with them. The NAI 2.8 release expands Nutanix Private Inference capabilities designed to deliver just that: improving performance while maximizing the value of existing hardware investments.

Fine-Tuning: tune smaller models (<8B parameters) with organization and domain-specific data to match the accuracy of large LLMs for targeted use cases, maximizing resource efficiency and controlling GPU spend.

Air-gapped NVIDIA NIM support: enable highly regulated customers with NVIDIA AI Enterprise (NVAIE) licenses to deploy the latest NVIDIA models via NVIDIA NIM microservices in air-gapped environments with operational simplicity.

Multi-Node & Multi-GPU Inference (Tech Preview): serve frontier-class models (100B+ parameters) with pipeline parallelism to run inference across multiple GPU nodes when single-node memory isn’t enough.

kvCache Offloading (Tech Preview): maximize GPU efficiency by moving KV cache context from GPU memory to CPU host memory when capacity is reached, minimizing redundant recomputations and enabling fast time-to-first-token (TTFT) during inference.

Extending Control and Governance in Nutanix Enterprise AI

Governing agent connectivity to enterprise tools is just one piece of the enterprise AI challenge. This release continues to strengthen the governance foundation of Nutanix Enterprise AI with platform-level capabilities designed to deliver greater governance with operational ease across both Nutanix Agent Gateway and Nutanix Private Inference. 

Custom Role Builder Support: empower organization admins to create custom roles with a catalog of 30+ permissions spanning user management, licensing, models, endpoints, API keys, and observability. Build roles from scratch or duplicate predefined roles to fit your organization’s unique access requirements.

See it in Action

NAI 2.8 is available now for all Nutanix Enterprise AI customers. Explore the latest capabilities and see how you can deliver greater governance, control, and efficiency to your enterprise AI deployments.

Check out our documentation page for more details or reach out to your Nutanix representative today.

*Tech Preview indicates these features should not be used in production environments.

1. Source: Gartner, Inc, Avoid Governance Mismatch: Classify AI Agents by Autonomy Level, Shiva Varma, 2 March 2026

©2026 Nutanix, Inc. All rights reserved. Nutanix, the Nutanix logo and all Nutanix product and service names mentioned are registered trademarks or trademarks of Nutanix, Inc. in the United States and other countries. GARTNER is a trademark of Gartner, Inc. and/or its affiliates. All other brand names mentioned are for identification purposes only and may be the trademarks of their respective holder(s). 

This content reflects an experiment in a test environment. Results, benefits, savings, or other outcomes described depend on a variety of factors including use case, individual requirements, and operating environments, and this publication should not be construed as a promise or obligation to deliver specific outcomes.

This content contains express and implied forward-looking statements, including but not limited to statements regarding our plans and expectations relating to new product features and technology in Tech Preview and under development, the capabilities of such product features and technology, and our plans to release product features and technology. Such statements are not historical facts and are instead based on our current expectations, estimates and beliefs. The development, release, and timing of any such products, features or functionalities are subject to change. Nutanix will not have any liability for any failure to deliver or delay in the delivery of any such products, features or functionalities. Any future product or product feature information is intended to outline general product directions, and is not a commitment, promise, or legal obligation for Nutanix to deliver any functionality. This information should not be used when making a purchasing decision.