AI First? Why CIOs Are Putting the Cart Before the Horse

The 2026 NetOps Reality: You Can't Modernize with AI Until You Fix Visibility and Automation
Jeremy Rossbach
Broadcom
Learn more about Broadcom

The conversation around AI in the enterprise has officially shifted from "if" to "how fast." But according to the State of Network Operations 2026 report from Broadcom, most organizations are unknowingly building their AI strategies on sand.

The data is clear: CIOs and network teams are putting the cart before the horse. AI cannot improve what the network cannot see, predict issues without historical context, automate processes that aren't standardized, or recommend fixes when the underlying telemetry is incomplete. If AI is the brain, then network observability is the nervous system that makes intelligent action possible.

The Core Problem: AI Needs Good Data, and Networks Aren't Providing It

The report reveals a troubling pattern:

  • 87% say cloud and internet usage creates critical blind spots.
  • 95% lack visibility into at least one major segment of their delivery chain.
  • 39% say insufficient visibility is directly impacting AI success.
  • 71% don't fully trust AI in network operations.

Why the mistrust?

Because AI is only as reliable as the data you feed it, and the data today is incomplete. Public cloud paths, remote work connections, ISP transport layers, and peering networks remain black boxes, yet companies are trying to deploy AI-driven triage and AI-driven automation on top of them. It's the equivalent of asking a self-driving car to navigate with half a windshield.

Before We Talk About AI, We Need to Talk About Order

There is a simple, non-negotiable sequence for modern networking success:

1. Visibility Enables Trust

You cannot automate what you cannot see, and you cannot trust AI-driven decisions when the data behind them is partial or ambiguous. The report makes this clear in no uncertain terms: cloud-to-cloud visibility, real-time path validation, ISP metrics, and public cloud telemetry are all required. Yet only 5% of respondents say they receive the ISP data they need, underscoring a simple truth that without visibility, trust is impossible.

2. Automation Enables Scale

Once you have reliable telemetry, patterns emerge: including baselines, repeated bottlenecks, behavior over time, capacity trends and known-good configurations. These patterns form the backbone of automation. However, today, only 27% of companies have mature automation practices. A full 70% are still early to mid-stage. Why so low? Because automation depends on accuracy. And accuracy depends on visibility. Automation cannot scale when your data is fragmented, when paths are unknown, or when you're still troubleshooting ISPs manually.

Which brings us to the final step.

3. AI Enables Foresight

AI is the multiplier. AI is the prediction engine. AI is the triage assistant that cuts MTTR and solves problems before users feel them. But AI is the last step and not the first. The report's adoption numbers prove enterprises are trying to skip ahead. 92% plan to use AI-enabled network operations solutions, but only 23% have anything deployed and a full 71% don't trust AI yet.

Why not? Because AI recommendations can't be trusted until the data is complete, the telemetry is continuous, the automation workflows are consistent and the visibility is fully end-to-end. AI doesn't magically fix missing data. AI makes missing data more dangerous.

The Right Path: Visibility → Automation → AI

Every CIO and NetOps leader should use this sequence as the modernization roadmap:

Step 1 - Expand Visibility - Goal: create trustworthy data

Expanding visibility means achieving full-path insight from the user all the way to the application, across both overlay and underlay networks. It requires comprehensive coverage of cloud environments, the public internet, and ISP infrastructure, combined with real-time and historical perspectives.

Step 2 - Scale with Automation - Goal: create consistent, repeatable execution

Take the high-fidelity telemetry and build: automated troubleshooting, automated remediation, automated policy deployment and automated performance baselining.

Step 3 - Apply AI for Foresight - Goal: create proactive, intelligent operations

Once trust and scale exist, AI becomes: predictive, accurate, reliable, safe and transformative.

But again — in that order.

The CIO Takeaway

AI is not your starting point; it is the reward for doing visibility and automation right. The message from the 2026 data is not subtle: AI initiatives are outpacing the network's readiness to support them. Until visibility gaps are closed and automation matures, network teams will continue to struggle, budgets will be strained, and enterprises will underperform on AI ROI. Your network observability stack is the AI data engine, your automation layer is the AI execution engine, and your AI platform is the intelligence engine. Mixing up this order doesn't accelerate progress; it sabotages it.

Final Thoughts

If enterprises want AI to truly deliver, powering autonomous triage, proactive remediation, intelligent routing, predictive detection, and real-time optimization, then the sequence is clear: visibility enables trust, automation enables scale, AI enables foresight and the companies that follow this order will win the AI era of NetOps.

Jeremy Rossbach is Chief Technical Evangelist - Network Observability at Broadcom

Related Links

Hot Topics

Related Vendors

The Latest

The Performance Bottleneck in Enterprise AI Has Been Exposed – Now We Must Fix It

Performance bottlenecks aren't uncommon when it comes to rolling out new technology, regardless of how capable or game-changing that technology might be. Every generation of new tech has encountered roadblocks that had to be overcome before it was truly able to shine. Virtualization forced organizations to rethink resource allocation, cloud transformation had us shift our focus toward scalability and elasticity, and microservices introduced entirely new challenges around observability and distributed systems. There's something different about AI, however ...

What Capital Markets Can Teach Enterprises About Integrated Data Infrastructure

Consider a single order represented across order-management, execution, and settlement systems. Each database, message broker, and application may be online and processing its own records correctly. Yet the workflow has failed if related events arrive on different clocks, rely on inconsistent state, or cannot be reconciled before an operational decision must be made ...

Why AI in IT Is Producing High Productivity but Heavy Workloads

AI now exists in almost every IT workflow. In a recent survey of more than 800 IT service professionals, all respondents indicated the use of AI in some form within their organization. But there's a growing paradox: if dashboards are clearing faster and alerts are resolved at unprecedented speed, why aren't IT service desks reporting lighter workloads? The research found that 71% of IT teams said their actual workload has remained flat or increased since adopting AI. This reality appears to contradict what we’ve been told about AI ...

Implementing AI Isn't the Hard Part Anymore - Adoption Is

Two years ago, almost every customer conversation about AI started with the same questions: Which model should we use? What can it do? Is it ready for the enterprise? Today, those discussions have moved on. CIOs are far more interested in how to govern AI, integrate it with existing systems, prepare their workforce and make it part of everyday operations. The challenge is no longer to prove that AI can deliver value. It's instead about how to embed AI into the business in a way that's secure, scalable and delivers measurable outcomes ...

The Incident Load Is Rising, but Recovery Times Aren't Moving

Two things happened to production incidents between 2023 and now, and they did not happen at the same speed. The first is that a class of dependency that barely existed three years ago now accounts for one incident in ten. Incidents disclosed by AI model and AI application providers rose from 1.7% of all disclosed unplanned incidents in 2023 to 10.7% in 2026 year to date, roughly a sixfold rise; that counts only incidents at AI companies themselves, so the true share is higher. The second is that the time to close an incident has not come down ...

Your AI Pipeline May Have a Document Observability Problem

When an AI assistant gives an incomplete or incorrect answer, teams often blame the model. They adjust prompts, switch models, increase context windows or test a new retrieval strategy. However the model may not be a problem. In many enterprise AI workflows, the problem begins inside the document-ingestion pipeline ...

The SIEM Isn't the Problem. Your Telemetry Architecture Is

If you talk to any security or observability teams right now, they're all fighting the same fire: their tooling was built to ingest X, but their sources are pumping Y and soon to be doing Z. The knee-jerk reaction is always the same: we need more platform. However, this reaction is wrong. Let me explain why, because the solution to this problem is foundational, not financial. Instead of hurling yet more money at the problem, make sure you've done what's needed upstream ...

As AI Scales Across Enterprises, Breaking Points Emerge

Rapid AI adoption and the unique ways AI workloads operate is redefining the scope and structure of what these teams must deliver. This shift is forcing organizations to rethink how they manage scale, automation, and control, according to The State of SRE and Platform Engineering 2026, a new report from Dynatrace ...

Preparing Network Infrastructure for AI Starts with the Basics

AI is usually talked about as a software tool, but it also depends heavily on the network behind it. Whether a company is using AI for chatbots, automation, monitoring, analytics, or employee support, all of that information has to move across the network in a reliable and secure way. That means AI is not just an application decision. It is also an infrastructure decision. Before organizations rush into AI, they should ask a simple question: Is our network ready to support it? ...

Agentic Operations Begin with Governed Middleware Access

Enterprise AI often lacks governed access to where business processes actually execute. Without that access, AI agents may be able to reason, but they cannot operate reliably across enterprise workflows. For AI agents to effectively carry out workflows, they will require integration-layer context and controls. Organizations can implement these prerequisites by providing AI with managed access to the middleware layer ...

AI First? Why CIOs Are Putting the Cart Before the Horse

The 2026 NetOps Reality: You Can't Modernize with AI Until You Fix Visibility and Automation
Jeremy Rossbach
Broadcom
Learn more about Broadcom

The conversation around AI in the enterprise has officially shifted from "if" to "how fast." But according to the State of Network Operations 2026 report from Broadcom, most organizations are unknowingly building their AI strategies on sand.

The data is clear: CIOs and network teams are putting the cart before the horse. AI cannot improve what the network cannot see, predict issues without historical context, automate processes that aren't standardized, or recommend fixes when the underlying telemetry is incomplete. If AI is the brain, then network observability is the nervous system that makes intelligent action possible.

The Core Problem: AI Needs Good Data, and Networks Aren't Providing It

The report reveals a troubling pattern:

  • 87% say cloud and internet usage creates critical blind spots.
  • 95% lack visibility into at least one major segment of their delivery chain.
  • 39% say insufficient visibility is directly impacting AI success.
  • 71% don't fully trust AI in network operations.

Why the mistrust?

Because AI is only as reliable as the data you feed it, and the data today is incomplete. Public cloud paths, remote work connections, ISP transport layers, and peering networks remain black boxes, yet companies are trying to deploy AI-driven triage and AI-driven automation on top of them. It's the equivalent of asking a self-driving car to navigate with half a windshield.

Before We Talk About AI, We Need to Talk About Order

There is a simple, non-negotiable sequence for modern networking success:

1. Visibility Enables Trust

You cannot automate what you cannot see, and you cannot trust AI-driven decisions when the data behind them is partial or ambiguous. The report makes this clear in no uncertain terms: cloud-to-cloud visibility, real-time path validation, ISP metrics, and public cloud telemetry are all required. Yet only 5% of respondents say they receive the ISP data they need, underscoring a simple truth that without visibility, trust is impossible.

2. Automation Enables Scale

Once you have reliable telemetry, patterns emerge: including baselines, repeated bottlenecks, behavior over time, capacity trends and known-good configurations. These patterns form the backbone of automation. However, today, only 27% of companies have mature automation practices. A full 70% are still early to mid-stage. Why so low? Because automation depends on accuracy. And accuracy depends on visibility. Automation cannot scale when your data is fragmented, when paths are unknown, or when you're still troubleshooting ISPs manually.

Which brings us to the final step.

3. AI Enables Foresight

AI is the multiplier. AI is the prediction engine. AI is the triage assistant that cuts MTTR and solves problems before users feel them. But AI is the last step and not the first. The report's adoption numbers prove enterprises are trying to skip ahead. 92% plan to use AI-enabled network operations solutions, but only 23% have anything deployed and a full 71% don't trust AI yet.

Why not? Because AI recommendations can't be trusted until the data is complete, the telemetry is continuous, the automation workflows are consistent and the visibility is fully end-to-end. AI doesn't magically fix missing data. AI makes missing data more dangerous.

The Right Path: Visibility → Automation → AI

Every CIO and NetOps leader should use this sequence as the modernization roadmap:

Step 1 - Expand Visibility - Goal: create trustworthy data

Expanding visibility means achieving full-path insight from the user all the way to the application, across both overlay and underlay networks. It requires comprehensive coverage of cloud environments, the public internet, and ISP infrastructure, combined with real-time and historical perspectives.

Step 2 - Scale with Automation - Goal: create consistent, repeatable execution

Take the high-fidelity telemetry and build: automated troubleshooting, automated remediation, automated policy deployment and automated performance baselining.

Step 3 - Apply AI for Foresight - Goal: create proactive, intelligent operations

Once trust and scale exist, AI becomes: predictive, accurate, reliable, safe and transformative.

But again — in that order.

The CIO Takeaway

AI is not your starting point; it is the reward for doing visibility and automation right. The message from the 2026 data is not subtle: AI initiatives are outpacing the network's readiness to support them. Until visibility gaps are closed and automation matures, network teams will continue to struggle, budgets will be strained, and enterprises will underperform on AI ROI. Your network observability stack is the AI data engine, your automation layer is the AI execution engine, and your AI platform is the intelligence engine. Mixing up this order doesn't accelerate progress; it sabotages it.

Final Thoughts

If enterprises want AI to truly deliver, powering autonomous triage, proactive remediation, intelligent routing, predictive detection, and real-time optimization, then the sequence is clear: visibility enables trust, automation enables scale, AI enables foresight and the companies that follow this order will win the AI era of NetOps.

Jeremy Rossbach is Chief Technical Evangelist - Network Observability at Broadcom

Related Links

Hot Topics

Related Vendors

The Latest

The Performance Bottleneck in Enterprise AI Has Been Exposed – Now We Must Fix It

Performance bottlenecks aren't uncommon when it comes to rolling out new technology, regardless of how capable or game-changing that technology might be. Every generation of new tech has encountered roadblocks that had to be overcome before it was truly able to shine. Virtualization forced organizations to rethink resource allocation, cloud transformation had us shift our focus toward scalability and elasticity, and microservices introduced entirely new challenges around observability and distributed systems. There's something different about AI, however ...

What Capital Markets Can Teach Enterprises About Integrated Data Infrastructure

Consider a single order represented across order-management, execution, and settlement systems. Each database, message broker, and application may be online and processing its own records correctly. Yet the workflow has failed if related events arrive on different clocks, rely on inconsistent state, or cannot be reconciled before an operational decision must be made ...

Why AI in IT Is Producing High Productivity but Heavy Workloads

AI now exists in almost every IT workflow. In a recent survey of more than 800 IT service professionals, all respondents indicated the use of AI in some form within their organization. But there's a growing paradox: if dashboards are clearing faster and alerts are resolved at unprecedented speed, why aren't IT service desks reporting lighter workloads? The research found that 71% of IT teams said their actual workload has remained flat or increased since adopting AI. This reality appears to contradict what we’ve been told about AI ...

Implementing AI Isn't the Hard Part Anymore - Adoption Is

Two years ago, almost every customer conversation about AI started with the same questions: Which model should we use? What can it do? Is it ready for the enterprise? Today, those discussions have moved on. CIOs are far more interested in how to govern AI, integrate it with existing systems, prepare their workforce and make it part of everyday operations. The challenge is no longer to prove that AI can deliver value. It's instead about how to embed AI into the business in a way that's secure, scalable and delivers measurable outcomes ...

The Incident Load Is Rising, but Recovery Times Aren't Moving

Two things happened to production incidents between 2023 and now, and they did not happen at the same speed. The first is that a class of dependency that barely existed three years ago now accounts for one incident in ten. Incidents disclosed by AI model and AI application providers rose from 1.7% of all disclosed unplanned incidents in 2023 to 10.7% in 2026 year to date, roughly a sixfold rise; that counts only incidents at AI companies themselves, so the true share is higher. The second is that the time to close an incident has not come down ...

Your AI Pipeline May Have a Document Observability Problem

When an AI assistant gives an incomplete or incorrect answer, teams often blame the model. They adjust prompts, switch models, increase context windows or test a new retrieval strategy. However the model may not be a problem. In many enterprise AI workflows, the problem begins inside the document-ingestion pipeline ...

The SIEM Isn't the Problem. Your Telemetry Architecture Is

If you talk to any security or observability teams right now, they're all fighting the same fire: their tooling was built to ingest X, but their sources are pumping Y and soon to be doing Z. The knee-jerk reaction is always the same: we need more platform. However, this reaction is wrong. Let me explain why, because the solution to this problem is foundational, not financial. Instead of hurling yet more money at the problem, make sure you've done what's needed upstream ...

As AI Scales Across Enterprises, Breaking Points Emerge

Rapid AI adoption and the unique ways AI workloads operate is redefining the scope and structure of what these teams must deliver. This shift is forcing organizations to rethink how they manage scale, automation, and control, according to The State of SRE and Platform Engineering 2026, a new report from Dynatrace ...

Preparing Network Infrastructure for AI Starts with the Basics

AI is usually talked about as a software tool, but it also depends heavily on the network behind it. Whether a company is using AI for chatbots, automation, monitoring, analytics, or employee support, all of that information has to move across the network in a reliable and secure way. That means AI is not just an application decision. It is also an infrastructure decision. Before organizations rush into AI, they should ask a simple question: Is our network ready to support it? ...

Agentic Operations Begin with Governed Middleware Access

Enterprise AI often lacks governed access to where business processes actually execute. Without that access, AI agents may be able to reason, but they cannot operate reliably across enterprise workflows. For AI agents to effectively carry out workflows, they will require integration-layer context and controls. Organizations can implement these prerequisites by providing AI with managed access to the middleware layer ...