The business case for AI in network operations | usagoldmines.com

AI has emerged as the priority for executives globally, with a staggering 97% of CEO’s pursuing investment in AI tools. This shows no signs of slowing down, as PwC projects that AI-related spending will inject $15.7 trillion into the global economy by 2030. This has sparked a push for operational efficiency as organizations – driven by competitive pressure – look to implement AI across all possible applications.

However, in network operations, the current approach to AI adoption has fallen short of expectations. This has led to growing skepticism, entering what Gartner calls a “trough of disillusionment” where technical, operational and ethical challenges are raising concerns about its long-term effectiveness.

The networking challenge for AI

Although traditional network management systems, enhanced by AI, handle tasks like intelligent traffic routing well, broader AI adoption in networking is hampered by significant challenges. For the typical operator, the network operations landscape tends to be a maze of legacy systems and fragmented data sources requiring consolidating and rationalizing, so inheriting an environment where AI solutions struggle to gain meaningful traction.

Despite AI’s impressive capacity to analyze large amounts of data, its difficulty in seamlessly integrating with existing network architectures limits its reliability for mission-critical operations. For example, in telecommunications, simple misconfigurations can take down vast networks and hit critical customer services particularly hard. In a sector that had a lot of growth when nascent operational solutions were being developed and data quality wasn’t always implemented, there is a lot of pause on the use of Broad AI solutions.

This mismatch suggests that conventional networking approaches and human oversight will remain essential components of network management for the foreseeable future.

Navigating operational roadblocks

Among the technical difficulties of AI adoption, organizations looking to develop their own applications continue to face significant operational and financial barriers to effective deployment. One bottleneck is talent, particularly in specialized development roles. This shortage has reached peak levels, with 60% of IT decision-makers identifying AI expertise as their most severe skills deficit – a gap that continues to widen as demand outpaces the available talent pool.

While larger organizations have more immediate access to the capital and in-house skills to build and train AI applications, smaller ones often lack the resources to invest in the necessary IT infrastructure, software and specialized skills, making AI adoption a considerable challenge.

Furthermore, many of the tangible gains AI will deliver, are largely internal. This means it’s increasingly difficult for organizations to get around the overhead that comes with deploying AI since they cannot reasonably pass this on to customers. Additionally, current use cases are not at the stage where they are delivering the widespread transformations people have come to associate with AI and typically deal with narrow AI applications solving backend processes like optimizing controlled automation routines and data flow efficiency.

As such, businesses need to carefully set their expectations as to whether AI will provide a measurable return on investment (ROI) before committing to widespread adoption.

The value of AI in network operations

It’s important to highlight that AI has demonstrated value in network operations, moving beyond basic network monitoring to become a tool for operators. By sifting through vast amounts of network data, AI systems uncover hidden patterns and potential trouble spots before they impact service. This helps teams shift from reactive to proactive optimization, ultimately delivering more reliable performance.

AI automation also plays an important role in taking on the heavy lifting of day-to-day operations. For example, intelligently optimizing data flow by ensuring efficient bandwidth allocation and traffic routing. Additionally, detecting anomalies and triggering predefined responses, reducing the need for manual intervention. By strategically deploying AI for specific tasks that complement current workflows, organizations can boost their network performance while maintaining human oversight of critical decisions.

Looking to the future, AI will be vital in shaping next-generation networks. As networking and cybersecurity become more closely linked, we are likely to see a market shift towards single-vendor frameworks. This is important because as we undergo this shift, AI’s natural ability to automate responses and provide predictive analysis will help converge network management under a single pane of glass.

As more adaptive networks become a necessity, this will help systems keep up with demand, predict failures before they occur and coordinate resources with minimal human oversight.

One step at a time

AI presents promising opportunities for network operations. While it excels at automating intricate processes, enhancing efficiency and anticipating issues, its real value lies in transforming the networks of the future.

However, Broad AI solutions utilizing full autonomy remains impractical due to AI’s current limitations in managing complex, high-risk scenarios from multiple disparate data sets. Additionally, the significant costs associated with delivering the right environment for AI to flourish calls for a carefully planned strategy.

The conversation shouldn’t be about AI replacing humans, it should be about giving them sophisticated tools that enhance their current capabilities. Therefore, we should see AI adoption as a gradual, long-term endeavor solving clear use cases rather than a rapid transformation.

There are some strong use cases now, but patience and intelligent planning will pave the way to broader transformation later down the line. Exercise patience, think practically and reap the benefits.

We’ve featured the best IT management tools.

This article was produced as part of TechRadarPro’s Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro