The overlooked role of networks in AI infrastructure – Unite.AI

Organizations are investing money in the GPUs, servers, models, and compute capabilities needed to bring AI into production. But agentic AI places different demands on the network.

A chatbot typically waits for someone to ask it a question, while an agent can be assigned a task, interact with different systems and applications, make decisions and take action. As agents take on more and more work within the company, much of that activity will happen without someone at the keyboard.

According to Google Cloud State of infrastructure in the age of artificial intelligence83% of organizations say they need infrastructure upgrades to support production-grade autonomous systems. And while computing power is part of the picture, organizations must also prioritize their network, because it will need to support tasks that don’t follow the typical model of a human employee.

The death of the 9-to-5 network

For decades, enterprise networks have been designed largely around how people work. Employees connect to applications, access data, and consume cloud services throughout the workday, creating relatively predictable traffic patterns that shape network architecture and operations.

Artificial intelligence and agent systems are changing this equation. Agents can perform tasks and interact with applications, APIs, data sources, and other agents at any time, creating more seamless and dynamic traffic flows across campus networks, data centers, cloud environments, and the WAN. As implementations grow, networks must support increasingly complex dependencies, making performance, latency, resiliency, and security critical to reliable execution.

These demands also highlight the limitations of traditional NetOps models, which rely on monitoring tools to identify problems and human operators to investigate and resolve them. The next generation of NetOps must go beyond managing individual devices and events to understand end-to-end application and agent dependencies, using deeper visibility, AI-driven correlation, root cause analysis, and intelligent automation to proactively optimize and resolve issues.

The opportunity is to move from networks managed with the assistance of AI to networks managed through AI-driven intelligence and governed autonomy. Network teams can spend less time troubleshooting repetitive problems and more time defining policies, establishing guardrails, and optimizing architectures. The death of the 9-to-5 isn’t simply about networks that run 24/7; This is a fundamental shift in what drives network demand and how networks need to be designed, secured and managed to support an always-on ecosystem of intelligent, autonomous systems.

From a few AI tools to fleets of agents

Most organizations are still at a stage where AI is easy to manage, with AI co-pilots, assistants and applications operating in different parts of the business.

However, IBM found that only 11% of leaders surveyed are fully prepared for the scale of AI agent deployment expected in the next 12 months. As implementations grow, network teams will have more tasks to understand and more problems to investigate, often without other people doing the work.

This makes it important to reduce the time engineers spend searching dashboards, gathering information, and escalating routine questions. Providing teams with relevant operational information and troubleshooting guidance when they need it can also facilitate the sharing of expertise across teams, rather than relying on a limited number of expert specialists to answer every question. People’s skills are better spent on decisions that require judgment, such as setting policies, setting limits, and assessing risk.

The internet needs a seat at the AI ​​strategy table

AI planning often starts with models, applications, data and computing resources, but networking should be part of those discussions from the start. Agents need the right context to provide actionable recommendations, and in network operations, that means understanding how information about users, devices, and applications fit together rather than looking at an alert or metric in isolation.

This context helps teams get to the source of a problem faster and decide what to do next. Human supervision is still important, especially as agents gain more autonomy, which requires well-defined limits on what they can do independently and controls to enforce those boundaries. Teams should also be able to see why a recommendation was made and what information it was based on before deciding whether to act on it.

Gartner predicts that by 2029, at least 70% of organizations using AI agents in their IT operations will have serious problems with material service, security, or costs, due at least in part to weak controls. That’s why visibility, automation, reliability and control must be built into AI infrastructure plans from the start.

Building infrastructure for an always-on AI company

Agentic AI changes the way enterprise networks are used. Business is no longer tied to when employees come online, and network teams will need to manage that change without allowing the work required to keep the network running to grow at the same rate.

Companies don’t need to completely overhaul their networks, but they should identify potential weaknesses in their current infrastructure as the use of AI grows, particularly where teams lack the visibility to quickly identify problems, where response times begin to suffer, or where routine troubleshooting still requires significant manual work.

A network built around the nine-to-five workday will have to give way to a network that supports people, applications and agents working side by side, often at very different times and speeds. Properly functioning infrastructure will ultimately result in the measures that businesses already care about: fewer outages, faster resolution when something goes wrong, better user experiences, and less time spent tracking down problems.