Agentic ITSM Tools Are Shipping Before Enterprises Are Ready
AI service management has moved beyond demos into production products. Vendors such as Ivanti and ServiceNow are already shipping agentic ITSM tools that can create incidents, submit requests, and search knowledge bases without human intervention. Ivanti’s autonomous service desk agent, for example, targets exactly the repetitive tickets that swamp many IT teams. ServiceNow leaders are likewise positioning agentic AI as a way to transform how IT services are delivered, shifting analysts toward higher‑value, proactive work. Yet enterprise IT readiness is lagging. Research cited by McKinsey shows most organisations still stuck in pilots, with only a small minority scaling agents in any function. That mismatch between vendor capability and buyer preparedness is becoming the central ITSM adoption barrier: tools are technically available, but the environments they land in were designed for ticket-passing humans, not autonomous agents executing decisions across systems.
The Plumbing Problem: Infrastructure and Workflow Gaps
Agentic ITSM tools depend on infrastructure that most organisations simply do not have. McKinsey’s work highlights that ticket-based, human-led workflows are structurally misaligned with agents that need to cross systems, call APIs, and act under policy controls. One multinational only achieved up to 80% automation of 450,000 annual tickets after rebuilding workflows and customer journeys specifically for agent-led resolution. In many estates, that redesign has not even started. Red Hat’s leadership has described enterprises as being forced "back to basics" on fundamentals like patching, underlining how brittle many environments are. For unified communications teams, the gap is very concrete: a single call-quality incident spans network telemetry, device health, carrier status, and ITSM records. If an agent cannot reliably traverse all of those, it simply bounces work back to humans, undermining the whole promise of AI service management and adding complexity instead of relief.
Governance, Observability, and Control: The Missing Operating Model
Even when infrastructure exists, most enterprises lack the governance and observability needed to run agentic ITSM safely at scale. Standard monitoring may confirm an AI service is online, but it rarely explains what an agent decided, which systems it changed, or why. Gartner analysts warn that this visibility gap makes scaling risky, because AI decisions are often opaque while their errors can cause real operational and reputational damage. Gartner also expects fewer than half of organisations deploying AI to have dedicated observability tooling in the near term. At the same time, vendors stress that systems must both detect issues and decide and act securely at scale. Bridging that gap requires a clear governance model: what agents are allowed to do, how policies are enforced, what gets logged, and how humans review and override automated actions. Without this, enterprise IT readiness remains theoretical and risk remains high.
Cost and Productivity Risks of Deploying Agents on Unprepared Estates
Preparedness gaps do more than slow down AI service management adoption; they actively create risk. McKinsey projects IT infrastructure costs will rise sharply as AI workloads grow, while budgets stay largely flat. In theory, agentic ITSM tools can relieve that pressure by reducing operating costs and redeploying service staff to higher‑value work, as seen in the multinational that automated most of its tickets and redeployed half its team. In practice, deploying agents into an unready estate has the opposite effect. Inference workloads accumulate without cost visibility, agents act on stale configuration data, and misfires generate more tickets instead of closing them. McKinsey identifies four prerequisites many IT and UC teams still lack: an accurate CMDB, API-exposed actions with embedded policy checks, a defined governance model, and active monitoring of inference costs. Until those are in place, promised productivity gains will remain elusive.
From Pilots to Scale: A Readiness Roadmap for Enterprise IT
Agentic ITSM tools are no longer a future concept; they are in buyers’ catalogs today. Yet McKinsey’s finding that 62% of organisations are still in pilot mode shows how far enterprise IT readiness must advance. The path forward is less about buying more AI and more about modernising the estate around it. That means treating an accurate CMDB, robust API layers, and cross-system workflow automation as strategic investments, not side projects. It also means designing service journeys for agent‑first resolution, with clear guardrails on when humans step in. Dedicated observability for AI decisions and costs should be built into early deployments, not added later. As ServiceNow’s and Ivanti’s agentic capabilities mature, the winners will be organisations that tackle this programme of work before procurement, so that when they switch on agents, those agents can operate with control, visibility, and measurable value.