Key Points:
- Despite high hopes, companies are not moving forward with advanced AI in their networks, stuck in early pilot projects for over 18 months.
- AI workloads are creating massive new demands for network bandwidth, especially in cloud and edge locations, straining existing infrastructure.
- Firms are turning to outside experts for help, with 81% increasing spending on managed service providers to overcome skills gaps and security worries.
Companies Stuck in Neutral on Enterprise AI for Networks, Report Finds
A new study from tech analysts IDC delivers a sobering message: companies aren’t making real progress with artificial intelligence (AI) in their computer networks, despite planning to do so for years. The 2026 IDC AI in Networking Special Report finds a growing gap between the high expectations for AI and the actual, stalled execution on the ground.
"For almost two years, organizations haven’t moved," said Mark Leary, research director at IDC. "The people who were at select use were still at select use. The people who were at substantial use were still at substantial use." This lack of movement comes as network teams face rising pressure from all sides. They are trying to use AI in two main ways: to support the heavy data needs of new AI applications, and to make network operations smarter and faster. In both areas, familiar roadblocks have brought advancement to a near standstill.
The top problem is security. "You have to fight AI with AI from a network security perspective," explained Brandon Butler, a senior research manager at IDC. Hackers are using AI too, making traditional defenses weaker. After security, the biggest hurdles are fitting new AI tools into old, complex network systems and finding enough skilled people to run them. "Most folks don’t feel their staff can fully evaluate and select the right solutions," Leary noted.
Because of these challenges, a huge 81% of organizations are now planning to spend more money on outside managed service providers (MSPs) for help with their AI projects. They are essentially hiring experts to bridge their internal knowledge and capability gaps.
Even while adoption lags, the need for better networks is exploding because of AI. The report shows that the pressure on network infrastructure is "already on." A massive 89% of data centers plan to boost their internal bandwidth by at least 11% in the next year, directly because of AI. The need for fast connections between data centers is also soaring, with 91% expecting similar growth. This is critical for companies using hybrid setups, which is the most common model. "The cloud is almost always involved," Leary said. "The biggest group mixes one cloud platform with one or more data centers."
For companies using Microsoft’s ecosystem, this is a direct call to action for Windows Server, Azure networking, and hybrid cloud strategies. The strain on inter-data center links and cloud connectivity highlights the need for robust, scalable solutions like Azure Virtual WAN or Windows Server’s software-defined networking (SDN) features. If a company’s network can’t handle the traffic from AI workloads running in Azure or on-premises Windows Servers, their entire AI initiative can fail.
The next big frontier is the "edge"—sites like factories, stores, or offices away from main data centers. Today, 27% of organizations are already running AI tasks at the edge, and over half (54%) plan to join them within two years. "Folks who are leveraging AI more extensively are already pushing workloads to the edge," Butler said. This will send network demands even higher, with edge bandwidth expected to grow by a huge 51% next year. Managing this mix of cloud, data center, and edge is a complex network challenge that aligns perfectly with Microsoft’s push for intelligent edge computing with Azure IoT and Windows for IoT.
Interestingly, the study found a surprising shift in how companies want to use AI. Nearly half (46%) now prefer AI systems that can make decisions and act on them automatically, without human approval each time. This shows a growing comfort with automation, driven by the fact that networks are too complicated to manage by hand and there aren’t enough experts to do the job.
At the same time, companies are losing faith in single, all-in-one "platform" solutions. They are instead picking "best-of-breed" tools that work better for specific jobs. The report suggests this is because early platforms didn’t deliver the promised simplicity and savings. Meanwhile, big cloud companies like Microsoft (with Azure) are becoming even more important as strategic partners in this space.
IDC advises network leaders to start small with clear, high-impact AI use cases—like automatically checking network settings or using AI to stop security threats faster. They should also strongly consider partnering with outside experts, the MSPs that companies are already turning to. "Avoiding a problem pays way more dividends than fixing one faster," Leary stated.
The bottom line is that the "widespread move toward managed services" shows companies know they can’t do everything alone. With network demands rising from AI, edge computing growing fast, and pressure for automated operations mounting, the next phase will be about turning plans into real, measurable results. "This isn’t about whether AI will reshape networking," Leary said. "It’s about how quickly organizations can adapt before the gap becomes too wide to close."
The ultimate goal for network leaders isn’t just faster tech. They believe AI can improve their whole business, with top hopes being better IT service levels (31%) and more operational efficiency (30%), not just cutting costs. For any business running on Microsoft infrastructure, getting the network ready for AI isn’t a tech side project—it’s a core business necessity for staying competitive in an AI-driven world.
Read the rest: Source Link
Don’t forget to check our list of Cheap Windows VPS Hosting providers, How to get Windows Server 2025, Try Windows 11 Pro for Workstations & browse Windows Azure content.
Remember to like our facebook and follow us on twitter @WindowsMode.