In modern enterprises, technology has become a direct driver of business operations. As the number of systems, users, and digital services continuously expands, IT teams face increasingly complex workflows, repetitive manual steps, and a surging volume of requests. CIOs are tasked with boosting operational efficiency without continually expanding headcount, while business departments expect fast, transparent services with less reliance on IT.
AI opens up a new paradigm by understanding context, analyzing data, offering recommendations, and directly participating in workflows. Combined with the Co-Creating Value, Complexity-Native, and AI-Native mindsets of ITIL 5, this demands a fundamental rethink of how enterprises design and operate IT Service Management (ITSM).
If AI is changing how people work, shouldn't how we operate ITSM change as well?
That is precisely the shift from IT Service Management to Digital Service Management.
ITSM was created to help IT deliver services in a structured, consistent, and controllable manner. Practices like Incident, Request, Problem, Change, Asset, and Configuration Management enabled enterprises to transform human-dependent operations into standardized, measurable processes.
That core value remains intact. However, the environment in which IT operates has fundamentally shifted.
A modern enterprise may operate hundreds of applications, multiple cloud platforms, continuous integration pipelines, thousands of devices, and countless digital services serving both internal employees and external customers. A single change in one system can ripple across many others. A seemingly simple request might require cross-departmental coordination among IT, Security, HR, Operations, and business units.
As complexity rises, an ITSM workflow heavily reliant on manual handoffs quickly becomes an operational bottleneck:
IT Teams spend significant time categorizing requests, hunting down information, checking statuses, routing tickets between teams, updating progress, and performing repetitive tasks. Disparate systems store fragmented data, forcing agents to manually stitch information together to understand the full picture.
CIOs and IT Directors face a scalability challenge. If service and user volumes continue to grow while workload scales linearly with headcount, the operating model becomes financially and operationally unsustainable.
Operations Managers struggle with visibility. When data is scattered across tools, obtaining a clear picture of service quality, operational bottlenecks, team workloads, or root causes behind missed SLAs becomes extremely difficult.
Service Users in Business Units have a straightforward expectation: they want fast service without needing to understand the underlying technical complexity.
This is why Service Management needs to be reimagined.
DSM is not simply expanding ITSM into a few non-IT departments. The true transformation lies in how enterprises define services and how those services are operated.
Traditional ITSM primarily focused on the question: How should IT deliver and support services?
DSM poses a broader question: How should digital services be designed, operated, and continuously improved to co-create value for the enterprise?
| Traditional ITSM | Modern DSM |
|---|---|
| IT-centric | Business & service-centric |
| Ticket-centric | Service-centric |
| Reactive | Proactive |
| Human-driven workflow | Human + AI workflow |
| Siloed data | Connected context |
| Manual handoffs | Intelligent orchestration |
| IT service | Digital service |
| Resolution-focused | Value-focused |
This distinction is particularly vital in the context of ITIL 5. While previous ITIL frameworks heavily emphasized standardizing and optimizing service delivery, the modern mindset centers on the co-creation of value. A service is no longer viewed as a static product handed over by IT to the business. Instead, value is co-created through the ongoing interaction among technology, people, processes, data, and service users. DSM serves as the practical vehicle for embedding this mindset into the enterprise operating model.
An enterprise might already be using ChatGPT, Copilot, or various AI tools, but that does not automatically mean AI has transformed its operations. Conversely, an enterprise with fewer AI tools that has embedded AI into critical workflows may actually be at a higher level of operational maturity.
This journey can be envisioned across 5 AI Maturity Levels:
At this stage, the enterprise recognizes AI's potential but remains focused on evaluating costs, security, compliance, and applicability. AI has not yet driven significant operational changes. A common question at this level is: "Is AI truly suitable for our organization?"
Focus: Identify real pain points and opportunities rather than rushing to deploy a long list of unverified use cases.
At Level 2, AI has become part of daily routines. Employees use AI to draft content, search for information, summarize documents, analyze data, or assist with individual tasks.
However, adoption remains largely ad-hoc and personal. Different employees use different tools without standard guidelines, official workflow integration, or measurable organizational impact.
Current State: AI is present within the enterprise, but it is not yet an organizational capability.
This represents the most critical tipping point.
AI moves from being an external personal utility to an integral part of the operational workflow itself. Systems can automatically summarize, categorize, clarify information, provide recommendations, or trigger appropriate actions.
From this point forward, enterprises begin measuring tangible outcomes - such as reduced resolution times, decreased manual workloads, improved SLA performance, and higher service quality.
Level 3 should serve as a realistic target for many enterprises in the initial phase of transformation. The bridge from Level 2 to Level 3 lies in workflow design, data quality, and operationalizing AI.
At this level, AI does not just assist individuals; it begins absorbing real workloads for service teams. AI autonomously handles defined sequences of tasks, coordinates multi-step actions, and escalates to humans only when judgment or formal approval is required.
At the highest level, AI is no longer implemented as isolated projects. It becomes an enterprise-wide capability active in analysis, decision-making, and operational refinement.
Here, AI does not merely help one team work faster - it fundamentally shapes how the enterprise designs services, allocates resources, and makes business decisions.
Not every organization needs to target Level 5 right away. An overly ambitious AI roadmap launched before processes, data, and governance are mature often results in a costly investment with low adoption.
Instead, enterprises should start with a simple question:
"Which maturity level fits our current capabilities and will deliver the clearest value over the next 12 months?"
For many organizations currently transitioning between Active and Operational, a pragmatic goal is moving AI from personal assistance into active operational workflows.
It is easy for an enterprise to brainstorm dozens of AI ideas: chatbots, incident prediction, knowledge generation, automated change, AI agents, and more. However, not all use cases deliver equal value or carry the same risk.
An effective approach is categorizing AI capabilities into 6 distinct groups: Creation, Curation, Clarification, Cognition, Communication, and Coordination.
This framework helps organizations avoid starting with "What cool things can AI do?" and instead ask: "What capability does AI need to build into our service workflow?"
This group covers use cases where AI creates new content based on existing context.
Examples: AI drafting incident communications, change plans, rollback plans, post-incident reports, or knowledge base articles.
Value: Creation is often an easy starting point because it significantly reduces content drafting time while keeping humans fully in control.
AI should not just generate more information; it must help make existing enterprise information reliable.
Examples: Grouping related alerts, detecting outdated knowledge, identifying duplicate tickets, or surfacing changes scattered across multiple data sources.
Value: While less flashy than chatbots or autonomous agents, Curation is foundational - data quality directly dictates the success of every downstream AI capability.
Clarification is an ideal starting point for many organizations.
Examples: AI summarizing lengthy ticket threads, translating technical jargon into plain language, restructuring fragmented info, or turning complex event chains into actionable context.
Instead of forcing an agent to spend minutes reading through ticket history, AI instantly answers: What is the issue? What has been tried? Which systems are involved? Who is affected? What is the recommended next step?
Value: The operational value of Clarification is realized quickly with relatively low risk, as humans can easily verify outputs before taking action.
At this stage, AI digs deeper into operational data to identify patterns, anomalies, and insights.
Examples: Identifying similar historical incidents, detecting operational anomalies, analyzing trends, or assisting in workload forecasting.
Prerequisite: Cognition requires clean operational data and sufficient historical depth. Without solid input data, enterprises should refrain from committing to complex KPIs based on AI predictions.
Once knowledge and service data reach maturity, AI can serve as a natural communication layer between users and services.
Users no longer need to figure out which form to fill or where the service catalogue lives. They simply describe their needs in natural language. AI understands intent, retrieves context, and directs the user to the correct service or workflow.
Value: This capability dramatically transforms the user experience, but its success hinges heavily on the quality and accuracy of underlying knowledge repositories.
This represents the clearest level of agentic AI. AI moves beyond understanding, answering, or suggesting - it begins orchestrating actions across workflows.
Examples: Identifying the required steps for a request, calling relevant API workflows, and monitoring execution through to completion.
Risk Control: Coordination carries the highest risk. Granting AI execution authority requires clear operational boundaries, strict data access controls, mandatory human approval checkpoints for high-risk actions, and robust rollback controls.
This is where many organizations stumble. Coordination - the agentic suite where AI is granted real execution power - is the most enticing group and often the first request from leadership. Yet it carries the highest risk and demands the highest data hygiene, governance, and organizational maturity.
The recommended implementation sequence based on the value-to-risk ratio:
Clarification First: Delivers the highest value-to-risk ratio, low activation cost, and immediate perceived value.
Curation Next: Establishes data quality for all remaining groups. Without strong Curation, Cognition and Coordination outputs will be untrustworthy.
Creation and Cognition: Introduced once data is clean and processes have stabilized.
Communication: Introduced once knowledge repositories are validated and approved.
Coordination Last: Implemented with four essential control gates:
Once you assess your organization's position on the AI maturity scale and use the 6C framework as a guiding map, start with a few well-defined workflows that suffer from clear pain points, possess relatively ready data, and offer measurable value. As initial use cases prove effective, expand into capabilities with higher autonomy and complexity.
This is the pragmatic path to transitioning from Modern ITSM to Modern DSM: not changing everything all at once, but systematically guiding AI from assisting humans → participating in workflows → absorbing workloads → becoming an integral part of your digital service operating model.
To quickly assess your current ITSM posture and locate your position on the AI maturity scale, contact the BiPlus team here!
