How Are Enterprises Approaching AI in Networking?

_{HOW ARE ENTERPRISES APPROACHING AI IN NETWORKING?}

^{AI-Driven Networks Are Already Widely Adopted}

What stage is your organization currently in regarding the adoption of AI-driven network solutions?

AI is already in widespread use in enterprise networking, and the majority (78.5%) of survey respondents have deployed AI-driven network solutions to some degree.

More than a quarter (27.8%) have fully integrated or autonomous operations, and it’s likely this number will grow as AI in networking continues to mature. Even more (28.9%) have implemented AI-driven network solutions across multiple areas.

Around a fifth (19.6%) are at the exploring/research, proof of concept or pilot stage. Only 2% have no current plans to introduce it.

_{HOW ARE ENTERPRISES APPROACHING AI IN NETWORKING?}

^{How are Enterprises Using AI?}

Given that many organizations are focusing on differentiating themselves through better customer experience, it’s no surprise that the most common AI use case in our survey was customer support or intelligent chatbots (52.4%). This was closely followed by security threat detection and response (51.8%), which we’ll explore in more detail later.

48.7% of respondents are using AI for network planning and capacity forecasting. Traditionally, capacity planning was based on historical records and patterns. These estimates struggled to predict unexpected network behavior and spikes in demand, leading to congestion, poor performance, and even outages.

This has been exacerbated by the rise of cloud, IoT, and AI, which have driven up traffic levels in ways that are often hard to predict. AI-powered solutions, on the other hand, use sophisticated algorithms that can predict capacity needs more accurately, improving network performance and cost-effectiveness.

In which areas are you currently using or piloting AI?

_{HOW ARE ENTERPRISES APPROACHING AI IN NETWORKING?}

^{Enterprises Prefer a Hybrid, Multi-Partner Approach to AI in Networking}

39.0% of respondents are approaching AI in network operations with a hybrid model across multiple partners. This might reflect a lack of confidence in any one provider to offer what enterprises need, but risks creating complexity that could impact adaptability and visibility.

A further 23.3% of respondents are using hyperscaler platforms, and just under a fifth (19.8%) are partnering with specialized managed service providers.

Which best describes your AI strategy for network operations?

_{HOW ARE ENTERPRISES APPROACHING AI IN NETWORKING?}

^{Security, Integration and Compliance Issues Are Holding Back AI Adoption}

Security, privacy and data quality concerns are the biggest barrier to AI adoption, with 23.8% of respondents reporting this as a challenge. Technical integration issues like legacy systems and vendor lock-in are a close second (20.4%), followed by regulatory and compliance uncertainty (15.3%).

When we asked our respondents to expand on their answer, several highlighted issues like incorrect or incomplete implementation, or adopting AI too fast. Others mentioned cultural issues like a lack of user adoption and employees fearing being replaced. Some discussed the fact that AI is a comparatively new technology that’s prone to making mistakes, meaning there’s uncertainty about how to use it, teething problems, and a lack of trust.

Which barrier will MOST slow down AI adoption in your organization?