AI adoption in networks is the norm despite its infancy

AI adoption is finally widespread

Different versions of AI have existed in networking for a long time. Jim Frey, principal analyst for networking at Enterprise Strategy Group and author of the report, listed a wide variety of AI types, including the following:

• Machine learning.
• Predictive AI.
• Causal AI.
• GenAI.

"[GenAI] is relatively more recent, but there are all these other sorts of related intelligent analytics that have been part of network management systems and the technology landscape for decades," Frey said. "We're seeing an interest in deploying GenAI and starting to use it more heavily."

As AI capabilities grow, it's no longer a question of whether enterprises will adopt AI into their networks but when. That time has finally arrived. Most survey respondents said their organizations are already deploying AI in their networks:

• Most organizations, around 42%, are in the mature stages of AI production, with active use across the enterprise.
• A slightly smaller percentage of organizations, 36%, are in the early stages of AI production.
• Lastly, 10% of respondents reported their organizations are not in the production stage of AI, but they have a pilot or proof-of-concept program.

The remaining respondents either had short- or long-term plans to adopt AI within their networks. While this data appears to show widescale AI adoption, excluding GenAI, Frey indicated possible confusion between the traditional and generative models of AI. Despite the confusion, Frey said the takeaway from this data is that enterprises adopt both AI types.

It should be unsurprising, then, that AI skills are now a necessary tool in any networking professional's toolkit. According to the survey, 75% of network professionals are very familiar with AI tools. A slightly smaller percentage -- 73% -- are very familiar with GenAI tools.

AI and GenAI skills are something enterprises now look for when hiring, too.

"[AI skills are] going to become something you need along with Word and PowerPoint," Frey said. "That's where we're headed, for sure."

GenAI adoption is moving quickly

Although the survey separates GenAI from other AI technologies, its adoption closely mirrors traditional AI adoption in networking:

• Most organizations, 42%, are in the mature stages of GenAI production.
• A little over a third of organizations, 35%, are in early production.
• Finally, 11% of organizations are in the pilot or proof-of-concept stages.

Despite GenAI's lower maturity rates compared to traditional AI, Frey said GenAI has already become mainstream in enterprise networks. He credited this to multiple forms of GenAI, such as ChatGPT, Google Gemini and Microsoft Copilot.

"There are multiple forms of generative AI, and we didn't ask which types [organizations] were using or how they're using it," Frey said. "Everybody's using it ... and it's becoming a standard part of everybody's toolkit."

GenAI deployment locations are diverse. Popular deployment locations include the cloud, SaaS, on-premises data centers and the network edge. Within these locations, GenAI proves to be a versatile tool with a range of use cases, including the following:

• ITOps.
• Cybersecurity or fraud.
• Data generation.
• Data insights.
• Employee productivity and tasks.

"There's still a lot of uncertainty about how to effectively deploy GenAI, even though we've had a couple of years with it in the marketplace," Frey said.

Still, uncertainty hasn't stopped organizations from deploying AI in their networks. Among the 12% of respondents whose organizations hadn't adopted AI yet, 8% plan to do so within 12 months, while 4% plan to do so in the next two years.

AI and GenAI are crucial to networking

Respondents agreed that AI is now critical for their enterprise networks. Enterprise Strategy Group's survey asked various questions regarding how critical AI has become to the network. The following are the combined percentages of respondents who answered either "strongly agree" or "agree:"

• Many respondents, 88%, said applying AI technology is a high priority for their business.
• Most respondents, 90%, believe networking is becoming more critical with the arrival of AI technologies.
• Similarly, 90% believe network management needs AI and automation.
• In addition, 86% of respondents believe AI will affect network operations.

Reflecting these results, both AI and GenAI highly affect planned, active and future network projects. By comparison, 51% of respondents said AI has a "high impact" on planned architecture projects, while 47% said the same about GenAI. In addition, 47% said AI and GenAI have a "high impact" on their active projects. Finally, 45% reported AI had a "high impact" on completed projects, while 44% reported the same about GenAI.

"Everybody's out there playing with it. Everyone's trying to figure out when to use it," Frey said. "Everyone I talked to, outside of the really big hyperscalers, is still in the early phases of trying to figure out exactly how to leverage this new technology because it's very expensive and very complex."

Enterprises are willing to accommodate the expense. IT budgets related to AI projects must increase, and most respondents expect a 5% to over 10% increase within the year. Over the next two years, respondents expect a similar increase in their budgets.

"It's interesting and significant in terms of the amount of budget increase that's being driven this way. It's more than we would typically see," Frey said.

Frey explained that, although not in the report, most of the conversations he's had indicated AI budgets and AI projects are driving network refreshes and upgrades. However, some respondents reported their budget for AI and GenAI will remain flat. Frey believes this is because these organizations have recently upgraded their networks and feel they can currently deploy AI.

Enterprise networks are large and complex. With so many different technologies and functions, AI has a different effect on each part of the network. The effect AI has on a network depends on what part it targets. This table compares the influence AI and GenAI will have on different types of networks.

While most organizations have adopted AI and GenAI technologies and agree both types are necessary for networking, that doesn't mean AI has reached full maturation. AI usage isn't fully realized, and it will take some time before the networking industry reaches a consensus on how to use it.

"There's still a lot of variability and change coming," Frey said. "So, that's going to be interesting to watch."

Nicole Viera is assistant site editor for Informa TechTarget's SearchNetworking site. She joined Informa TechTarget as an editor and writer in 2024.