New EMA study reveals gaps in AI adoption in network operations

A survey of 458 IT professionals found they are actively working on AI-driven NetOps, but only 35% reported complete success

Enterprise Management Associates (EMA™), a leading IT research and consulting firm, announced a new research report, “AI-Driven NetOps: How Enterprises Are Adopting Intelligent Network Management Solutions,” written by Shamus McGillicuddy, EMA’s Vice President of Network Infrastructure and Operations Research.

This study examines how enterprises are implementing AI-driven network management solutions, including both commercial products and internally developed tools. While this study shows that organizations are highly leveraging AI for network operations, it also reveals significant gaps in execution. Only 35% of respondents believe their AI-powered network management efforts are completely successful.

“The quality of network data is the deciding factor for AI,” McGillicuddy says. “The success of AI-driven networking is highly correlated to confidence in data quality. Today’s network operations are plagued by a variety of data issues, including data collection errors, insufficient documentation, and proprietary data formats. IT organizations must clean up their network data before investing in AI.”

Since 2023, EMA’s Network Infrastructure and Operations practice has closely tracked the evolution of AI in network management. The AI ​​revolution occurred during this period. Large-scale language models (LLMs) and AI agents have become mainstream, and an investment gold rush has begun. In the IT industry, many vendors are doubling down on their investments in AI, moving beyond machine learning and leveraging LLM and agents to develop advanced AI-driven IT operations capabilities, bringing the industry closer than ever to autonomous operations. In the world of network infrastructure and operations, leading vendors are recognizing AI technology as a key focus for innovation and differentiation. Additionally, EMA analysts are observing a shift in perspective among network engineers and architects. Five years ago, they were skeptical about the value of AI-driven network management, but now they expect AI solutions from strategic vendors.

Also read: AiThority Interview Featuring: Pranav Nambiar, Senior Vice President, DigitalOcean AI/ML and PaaS

This study highlights several challenges that limit the success of AI in NetOps today. Only 39% of respondents said they had complete confidence in their organization’s ability to evaluate AI-driven network management solutions, and only 44% expressed complete confidence in the quality of network data supporting AI initiatives. Despite these concerns, adoption continues to accelerate, with 59% of organizations using AI capabilities provided by network management vendors and 52% using their own IT and security data to train AI models.

The study investigated:

• How enterprises are approaching AI-driven network management solutions
• Which AI technologies are being adopted across NetOps teams?
• How organizations are applying AI, the challenges they’re facing, and the benefits they’re realizing.
• New best practices for successful AI deployment in network operations

EMA’s goal is to provide IT organizations with a snapshot of the current state of AI-driven network management and what they hope for in the future.

This independent research is sponsored by BlueCat, Broadcom, EfficientIP, Gigamon, IBM, Infoblox, and Riverbed.