Back to Resources

The current state of AI in the ITSM industry 

Around June 2024, we set out to assess the state of AI in the ITSM industry. We conducted a survey and developed a research report, uncovering several key insights. While the data is now over six months old, which is a considerable period in the rapidly evolving AI landscape, the trends identified are likely even more pronounced today. As AI adoption accelerates across both consumer and enterprise products, it’s reasonable to assume that the shifts we observed have only gained momentum.

Here are the few key numbers that stood out:

• 81% of respondents cited process streamlining and cost reduction as the primary drivers for AI adoption. IT, traditionally viewed as a cost center, has always prioritized solutions that optimize both cost and efficiency.

• 51% predict that within the next five years, 41%–100% of ITSM processes will be AI-driven.

• 81% also believe that AI in ITSM will lead to more consistent service delivery.

These findings strongly suggest that AI is no longer just a differentiator in the ITSM market—it’s becoming a fundamental necessity for service delivery operations.

However, despite this growing importance, many organizations are still unsure how to effectively integrate AI into their existing ITSM frameworks. Our survey reflected this uncertainty: 71% of organizations are still in the research or pilot stage of AI adoption for IT support and ITSM operations.

With this notion, we set out to demystify how AI can reinforce and enhance traditional ITSM frameworks, addressing the common service delivery challenges with AI-powered capabilities.

 Weaving AI to enhance traditional ITSM workflows 

We identified five common challenges that organizations face across core ITSM practices. For this blog, I’ll focus on two key cases to illustrate our approach to addressing these challenges. This blog gives you a glimpse, but I’ll be breaking down all five use cases in my keynote at SDI Spark ’25. Be sure to find me on the main stage at 10:30am on March 28!

The two extremes of incident management

Every organization pays its dues to IT incidents, from minor disruptions to major outages. We took a real-world scenario where a retail chain faced challenges across the incident severity spectrum.

On one end, the company was overwhelmed with L1 incidents, such as locked user accounts, VPN failures, and laptop malfunctions. At the same time, a major outage brought down its entire point-of-sale system, crippling operations.

With IT teams stretched thin, pinpointing the root cause became an uphill battle. Was it a cyberattack? An insider threat? ISP issues? A third-party outage? The process involved sifting through endless logs and lengthy debates.

However, with the right AI-driven approach, resolution times for both L1 and major incidents can be drastically reduced:

• For L1 incidents, instead of relying on chatbots that simply suggest knowledge articles, organizations—depending on their IT maturity—can leverage user behaviour monitoring and ML models to predict, prevent, and resolve issues before they even reach the service desk.

• For major incidents, AI can accelerate root cause analysis through intelligent workflows, custom LLMs, and predictive ML models, eliminating the guesswork and enabling faster recovery.

The curious case of a travel service request

This next scenario has become increasingly common in the post-pandemic world.

A developer working for a remote Singapore-based organization plans to attend an industry conference in Las Vegas. They submit a travel request, receive approval from their manager, book a flight, and head to Vegas.

Midway through the conference, their corporate VPN fails, leaving them unable to access work resources. In a rush, they download a free VPN—only to receive a chilling email:

“Your systems are now encrypted.”

While the security breach itself is a serious concern, the root of the problem began much earlier—at the service request stage. IT was completely out of the loop, and the developer’s laptop was never properly encrypted.

Surprisingly, AI can play a crucial role in preventing such incidents.

• Traditionally, service requests are assigned using predictive AI models, while LLMs assist with VPN setup and troubleshooting. But with advanced AI capabilities, service delivery can be hyper-personalized to proactively prevent failures.

• By integrating case extraction, AI agents, and recommendations based on historical corporate data, organizations can build a foolproof system—ensuring critical lapses like VPN failures and encryption gaps never happen in the first place.

 Join Me at SDI Spark 2025 

I love sharing insights with the ITSM community, but for now, I’m keeping things a bit mysterious—so we can meet in person and dive into these use cases together.

In my keynote session, I’ll take a deep dive into five ITSM use cases across modules including incident, service request, change, asset, and knowledge management. Using workflow frameworks, I’ll showcase how the right AI technologies can transform these areas for efficiency, automation, and resilience.

As a bonus, I’ll also outline a high-level roadmap for organizations looking to kick-start their AI journey in ITSM.

Let’s connect at Spark ’25. I will be on the main stage at 10:30am on March 28! See you there! – https://www.sdiconference.co.uk/

Ashwin Ram Ragupathi
ITSM Marketing Manager, ManageEngine

About the author:

Ashwin is a seasoned ITSM expert with over a decade of experience in generating best practice articles, blogs, and product collaterals. He has also authored articles and delivered webinars on AI in ITSM, including the ebook “The AI Advantage: Use Cases and Scenarios on How AI Will Redefine the Way IT Service Desks Work.”

Ashwin is a regular presenter at the ManageEngine UserConferences across the globe, delivering sessions that help customers leverage ManageEngine solutions to address key IT challenges. He has also hosted fireside chats with industry leaders, and CXOs building engaging discussions on the pressing tech challenges.