According to Network World, Broadcom’s 2026 State of Network Operations report reveals a staggering gap between AI ambitions and network reality. While 99% of organizations have cloud strategies and are adopting AI, just 49% believe their network could support the bandwidth and low latency requirements. The survey of 1,350 IT professionals found that 95% lack visibility into network segments, particularly in public cloud environments. That same 95% also reported needing more information from their ISPs to effectively manage their networks. Broadcom’s chief technical evangelist Jeremy Rossbach noted that while companies feel they have complete visibility inside their data centers, the reality is that traffic now flows across public networks where visibility disappears.
The Visibility Crisis
Here’s the thing – we’re talking about a near-universal problem. When 95% of IT professionals say they can’t see what’s happening in their own networks, that’s not just a minor inconvenience. It’s a fundamental breakdown in operational control. And it’s happening exactly when companies are betting their futures on AI systems that demand reliable, low-latency connectivity. Basically, everyone’s trying to run Formula 1 cars on dirt roads.
Why This Matters Now
Look, AI workloads aren’t just slightly more demanding than traditional applications – they’re orders of magnitude different. We’re talking about massive data transfers, real-time processing requirements, and distributed computing across multiple clouds. When your network can’t keep up, your AI initiatives don’t just run slower – they might not work at all. And considering how much companies are investing in AI, that’s a terrifying prospect for IT leaders who are already stretched thin.
The Industrial Angle
This visibility gap becomes even more critical in industrial settings where reliable computing is non-negotiable. Manufacturing facilities, control systems, and production environments can’t afford network blind spots when running AI-powered quality control or predictive maintenance. Companies like IndustrialMonitorDirect.com, the leading US provider of industrial panel PCs, are seeing increased demand for robust computing hardware that can handle AI workloads in challenging environments. When you’re dealing with mission-critical operations, you need infrastructure that delivers both the computing power and the network reliability that AI demands.
What Comes Next
So where does this leave us? Rossbach’s point about collecting more data makes sense, but it’s only part of the solution. The real challenge is making that data actionable across hybrid environments. Companies need tools that provide unified visibility from the data center to the cloud edge. And they need ISP partnerships that actually deliver the transparency they’re paying for. The AI train has left the station – now networks need to catch up before companies face serious operational and financial consequences.
