Share

Many organizations are frustrated that investments in AI have not delivered anticipated results, especially regarding IT Service Management (ITSM).

Introducing AI into an ITSM environment offers huge potential—but most organizations encounter similar obstacles. Here are some of the common challenges that organizations encounter:

• Data and process maturity problems – AI amplifies immaturity—it doesn’t fix it.
• Organizational and cultural resistance -Fear and misunderstanding slow adoption.
• Lack of structured governance and continuous improvement – AI needs ongoing oversight and continual tuning; it is not a one-time deployment.
• Unrealistic Expectations – Senior management often expects AI to immediately reduce headcount, solve all categorization and routing issues, replace Tier 1 support and deliver human-level accuracy…. All on day one. AI is a tool; it is not magic.

But despite these well-publicized challenges, many organizations jump in headfirst and invest in AI tools that promise dramatic improvement of their ITSM capabilities. But the excitement of implementing that AI-enabled chatbot or copilot quickly fades when the organization starts to realize that their ITSM environment just isn’t ready.

Just like you can’t build a house on a shaky foundation, implementing an effective and impactful approach to AI adoption within ITSM must start with a strong foundation. Good ITSM is part of that strong foundation.

What is “Good AI”?

Let’s first discuss “Good AI”. Good AI is:

• Aligned with human goals and values – Good AI supports human intentions, helping people make better decisions, work more efficiently, or access information. Good AI does not mislead, manipulate, or undermine autonomy.
• Safe, predictable, and reliable – Good AI behaves consistently and avoids unexpected or harmful outcomes. It should be transparent regarding its limitations.
• Fair and unbiased – Minimizes bias and treats groups and individuals equitably.
• Respects privacy – protects user data, uses the minimum data necessary and allows users control over how data is collected and used.
• Transparent and explainable – People should be able to understand how AI reached a conclusion, what data it used to reach that conclusion, and how confident it is in its result.
• Human-centered – Designed for real human needs, intuitive/easy to interact with, and supportive (not replacing humans where judgement and empathy are vital.

But “Good AI” cannot fix “Bad ITSM”.

What is “Good ITSM”?

As I’ve said before, “good ITSM” is a business enabler, not just an IT function.  Good ITSM:

• Standardizes how work gets done across the organization, improving both productivity and throughput.
• Provides clarity and transparency into how value flows through the organization, enabling better visibility of end‑to‑end value streams.
• Identifies and defines services and processes that underpin organizational value streams, helping connect IT work directly to business outcomes.
• Brings repeatability, reliability, and measurability to all aspects of the organization—not just IT operations.
• Follows a human-centric approach emphasizing outcomes, experience, and putting people at the center of service delivery.

Good ITSM is foundational for realizing the benefits of good AI. Unfortunately, many organizations wanting to implement AI follow the same mistaken approach to technology adoption…by starting with the technology, and “working backwards” to process and people. These organizations are trying to build their AI houses on shaky ground – a recipe for unnecessary cost, frustration, and missed expectations and opportunities.

“But we already bought the AI tool”

Perhaps your organization is on this same path to AI adoption – starting with the technology. And likely, your organization is encountering the same challenges that many other organizations face when introducing AI to their ITSM environments. It may seem like you’re facing a “mission impossible”. What steps can be taken to turn this difficult circumstance into a positive outcome?

How about using AI adoption as a catalyst for shoring up the shaky ground of bad ITSM? Here’s a suggested approach:

• Establish a goal – Define and agree on two to three business-oriented goals for AI adoption within ITSM. AI within ITSM should not be approached as a point solution; rather, AI should be considered within the broader perspective of ITSM. Defining overarching goals for AI in ITSM – in business terms – ensures that broader perspective.
• Define how those goals will be measured – Measures should be more than just from an IT operations perspective. Measures should also reflect things like reduction of friction, accuracy of the AI-provided solution, user satisfaction, and usability.
• Pick a value stream – Start with one high-impact, highly visible value stream that would benefit both the IT consumer and the IT organization, like “onboard a new employee”, or “provide access”. Identify and review the procedures that support the steps identified in the value stream. Are those procedures documented and in use? Are there steps that are missing procedures?
• Involve the right people – As with most value streams, success depends on the engagement and contributions of people both within and outside of the IT team.
• Conduct experiments – It is impractical to think that you’ll be able to accomplish the goals and objectives of AI adoption within ITSM in a single go. Use small Plan-Do-Check-Act (PDCA) loops to experiment and learn, not only about the AI tool’s capabilities, but also how best to leverage those capabilities within your ITSM environment. PDCA loops will also help identify where ITSM practices just aren’t ready: poor data capture or quality, dead-end workflows, inconsistencies in categorization and escalation, and other “bad ITSM” behaviors.
• Share the learning – Sharing the learning improves collaboration, strengthens understanding of both ITSM and AI, and helps build momentum for further adoption and use.

Ready to turn AI frustration into AI advantage?

AI won’t rescue bad ITSM—it will expose it. The organizations that succeed with AI aren’t the ones chasing shiny technology; they’re the ones with strong processes, clear value streams, and a culture ready for change. When ITSM is solid, AI becomes a multiplier. When it’s not, AI becomes a mirror.

Use this moment. Let AI adoption be the spark that fixes shaky workflows, strengthens data quality, and aligns teams around meaningful outcomes. Start small, learn fast, and build momentum.

Ready to turn AI frustration into AI advantage? Start by strengthening your ITSM foundation—because that’s where real transformation begins.

Share