Automation & AI Articles

A quarter of UK enterprise AI agent deployments aren’t paying back, and new research from KTSL and BMC Helix shows the reasons have almost nothing to do with the technology. This article looks at what’s going wrong inside deployments, why model choice matters less than most vendors suggest, and what enterprises would need to change to close the return gap.
AI agents are automating not just operational IT tasks but parts of ITSM platform implementation, configuration, and maintenance. Richard Mendis argues this shift is underway, and that as it matures, organizations should expect lower costs, less manual intervention, and systems that manage themselves. A look at where self-healing ITSM is credible today, and where vendor claims still outrun reality.
Mathies Wähner on why Agentic AI fails in ITSM. Drop it into an operating model that isn’t ready and it doesn’t make you smarter, just faster at being wrong, while quietly removing the people who used to catch the mistakes. The real question isn’t how to implement Agentic AI, but whether your operating model can survive it once it’s running at scale.
Mathies Wähner on why Agentic AI fails in ITSM. Drop it into an operating model that isn’t ready and it doesn’t make you smarter, just faster at being wrong, while quietly removing the people who used to catch the mistakes. The real question isn’t how to implement Agentic AI, but whether your operating model can survive it once it’s running at scale.
Somebody announces the death of service management every few years, and they’ve been doing it for 35 years now. Barclay Rae has heard every version and thinks they all miss the same thing: the job at the heart of service management is human, not technological. That’s why no new wave of tech has managed to kill it, and AI has made the case for it stronger.
Every new tool promises to simplify IT operations and adds another layer of complexity instead. Rui Alves argues bolting AI onto disconnected systems just gives you automation without intelligence, and the CTOs getting real value in 2026 are doing the unglamorous work first: connecting service management, monitoring, assets, FinOps, and governance into a single operational layer.
Every few years, the IT industry settles on a new savior. Agile. Then DevOps. Now AI. Kaimar Karu argues organizations adopt each one without first knowing what they want from it, then act surprised when results don’t live up to the hype. In this episode of Roman Jouravlev’s Conversations with Giants series, Kaimar covers why AI should never be the goal in itself and the skills technology won’t replace.
Read any ITSM platform brochure today and the same cluster of AI terms stares back: agentic AI, AI agents, RAG, AIOps, MCP. The vocabulary is moving faster than most IT teams can keep up with. Raghav S of ManageEngine explains what each one means in an ITSM context, how they differ, and how to tell which capability fits a problem you have.
Value is what the customer decides it is, not what the IT department or a framework says. It’s the test Stuart Rance has applied across a thirty-year career: would anyone on the receiving end say your work created value for them? In episode three of Conversations with Giants he explains why he hates tool replacement projects, and why using agentic AI to cut headcount is the wrong use of the technology.
If AI is the strategic priority, why are the teams meant to implement it still buried in ticket queues? John Mathieu of Allari argues the blocker isn’t culture, skills, or tooling but the operating model itself: reactive work and AI projects compete for the same people, and reactive always wins. His fix is bifurcated execution, separating the two into distinct streams with protected capacity.
It’s 9am and a sales manager hits a VPN error before a client call. Rather than wait on a service desk ticket, she pastes the error into an AI tool and gets a fix in seconds. Problem solved, but no incident recorded. Judin Joan Soundarya S of ManageEngine looks at what shadow AI costs problem management and security operations, and how ITSM platforms can regain visibility.
Everyone is selling agentic AI, but the gap between the weakest and strongest versions is enormous. Manish Sharma of Rezolve.ai sets out a four-stage maturity model for AI in ITSM, from legacy retrieval through reactive assistants and process agents to true agentic systems that reason, act across connected systems, and catch problems nobody asked. He also gives you the questions to put to any vendor claiming agentic AI.
For over 20 years, ITSM has repeated the same failures. Through the ABC cards, Paul Wilkinson has spent decades tracing how culture, leadership, and behavior consistently derail success, not tools or frameworks. AI is now the latest shiny thing being sold to organizations that haven’t fixed the underlying pattern. This article looks at why the industry still hasn’t learned, and what it would take.
The mismatch between producer effort and user value is something Mark Smalley, IT Paradigmologist and author of nine books on digital and service management, has been thinking about for decades. In this article on the first episode of Roman Jouravlev’s Conversations with Giants series, Mark covers several ideas that have held up across four decades in ITSM.
AI is already here. The level of AI inclusion in ITSM tools and platforms is remarkable when you look at it closely. The question isn’t whether to engage with AI, it’s whether your organization has the right foundations to make it work. Trust, governance, and a clear reason for adopting it in the first place separate the organizations getting results from those that aren’t.