Sign up for the ITPro Today newsletter
Stay on top of the IT universe with commentary, news analysis, how-to's, and tips delivered to your inbox daily.
Generative AI's Impact on ITSM: Is It a Game-Changer for IT Engineers?
Despite its potential, generative AI's impact on overall productivity and efficiency in ITSM is likely to be modest. Here's why.
5 Min Read
Alamy
Can generative AI improve IT service management (ITSM)? Definitely. Will it make your IT engineers 10 times more productive? Probably not.
That's the short version of the impact that generative AI is likely to have on ITSM. For the longer version, keep reading as we unpack the value that genAI brings to ITSM, as well as the limitations it faces.
ITSM, short for IT service management, is the set of processes that an organization uses to deliver IT services. ITSM covers all elements of IT — from ticketing and help desk services that support end users, to server infrastructure management, to application deployment and beyond.
Because virtually all parts of modern organizations rely on IT services in one way or another, ITSM is one of the foundational practices that helps businesses operate smoothly and efficiently.
Traditionally, IT engineers carried out most ITSM processes manually. They managed tickets, deployed applications, and so on by hand. AI-based analytics tools might have helped them parse information in some cases, but most of the work of handling incidents or deploying new resources required manual effort.
But generative AI — meaning AI tools and services that generate new content — has the potential to empower IT teams to automate many other facets of their ITSM processes, such as:
Ticket summarization: Generative AI tools could automatically parse the information submitted by users when they submit support tickets, then summarize or refine it to make the tickets easier for IT engineers to read.
Alert summarization: Similarly, generative AI can summarize alerts created by monitoring systems to help IT engineers pick out the information that is most important.
Generating end-user communications: When IT engineers interface with end users — for example, when they send emails in response to tickets or support requests — they could rely on generative AI to draft the content for them.
Configuring IT resources: Generative AI tools could produce configurations for applications, servers, and other resources that IT teams need to deploy. This is especially true for teams that manage configurations using code, which is easy for generative AI technology to produce and deploy.
Documentation generation: By automatically producing content for knowledge bases, generative AI could help build out documentation and eliminate the need for engineers to perform this work by hand.
In these ways and more, generative AI has real potential to speed ITSM processes, as well as eliminate or reduce the time IT engineers spend on toilsome tasks that they don't enjoy. Indeed, it's already doing many of these things, thanks to genAI features that ITSM vendors have integrated into their products.
That said, the value that generative AI can ultimately offer to ITSM is likely to be limited, for two reasons.
One reason is that, for the most part, the productivity and efficiency gains that generative AI can bring to ITSM are marginal. For example, although the ability to summarize tickets might increase the rate at which IT engineers can review tickets by perhaps 30% or 40%, its overall impact on the entire IT ticketing process is lower. Why? Because reading tickets is only one step in ticket management. Someone still must decide how to respond to the ticket, carry out the response, review the results, and then close the ticket. Generative AI can help only with some of those tasks.
At the end of the day, you might be able to close out tickets 10% faster thanks to generative AI — which is a considerable advantage but not a revolutionary one.
Likewise, consider the process of creating documentation. Generative AI might be able to draft a knowledge base article for an IT process or tool in just a few minutes, whereas it could take an engineer half a day to do the same work by hand. That's great, but because creating documentation is typically a small part of an IT engineer's job, the ability to speed up documentation management is likely to have a small overall impact on the productivity of a given engineer.
The second major challenge of using generative AI in ITSM is the risk that genAI tools will "hallucinate," leading them to produce inaccurate or unreliable information.
Because of hallucination risks, it will be impossible for most IT teams to trust genAI-produced content blindly. They'll always need someone to review automatically generated configurations, documentation, end-user communications, and so on. This challenge reduces even further the productivity gains that genAI stands to offer because it means that even in the case of ITSM processes that genAI can speed up dramatically, someone will have to step in and perform the manual, time-consuming work necessary to verify the accuracy of whatever the AI produced.
This would be less of an issue if there were a large tolerance for error in ITSM processes. But there's not. Even if generative AI services make stuff up only 1% of the time, a 1% error rate is a big problem if you're deploying mission-critical infrastructure or documenting how an important IT service works.
There are plenty of reasons for IT teams to embrace ITSM tools and platforms that take advantage of generative AI to accelerate processes. But it's critical not to overestimate the value that generative AI stands to offer in this context. Expect overall productivity and efficiency gains in the low double-digits, at best, and don't expect AI to automate any mission-critical IT processes entirely. You'll always need a human in the loop.
About the Author(s)
You May Also Like
