SkarpChapter 17 of 19
ITIL and AI: Responsible, Value-Focused Automation
Connect ITIL’s principles and value system to modern AI use cases, examining how automation, data, and ethics shape service management in the AI and cloud era.
AI in ITIL Today: Setting the Scene
AI as Enabler, Not a New Process
In ITIL 4 (basis for ITIL 5 Foundation), AI is not a separate process. It is a set of technologies that support service management and help organizations deliver and improve services.
Two Anchors: SVS and Principles
To place AI in ITIL, anchor on: 1) the service value system (SVS), which shows how all parts work together to enable value, and 2) the guiding principles, especially optimize and automate.
Why AI Matters in 2026
Modern services are digital, cloud-based, and data-rich. AI supports automation, data-driven decisions, and better user experiences such as 24/7 virtual agents and personalized portals.
What You Need for the Exam
You do not need to build AI models. You must explain how AI supports value, spot examples that follow or break guiding principles, and apply continual improvement thinking to AI scenarios.
Core AI Concepts for Service Management
Key AI Types in ITIL Context
Common AI types: machine learning (patterns, predictions), NLP (language), generative AI (new content), predictive analytics (forecasts), and automation/orchestration (execute actions).
AI + Automation
AI often decides what and when; automation tools do the how. For example, an ML model flags a risky change, and an orchestration tool blocks deployment.
Value Co-Creation Lens
ITIL focuses on how AI supports value co-creation: joint activities of provider and consumer. Ask: which outcomes are improved, and how are utility and warranty affected?
Exam-Relevant Focus
You are not tested on AI math. You are tested on explaining how AI supports outcomes, improves service management practices, and aligns with ITIL principles.
AI in the Service Value System and Value Chain
AI Across the Value Chain
The service value chain transforms demand into value. AI can support Plan, Improve, Engage, Design & transition, Obtain/build, and Deliver & support activities.
Plan and Improve with AI
Plan: AI forecasts demand and capacity. Improve: AI mines tickets and feedback to find patterns and suggest high-impact improvements.
Engage, Design, Build, Support
Engage: chatbots and virtual agents. Design & transition: risk prediction for changes. Obtain/build: AI-assisted coding and testing. Deliver & support: AIOps and automated remediation.
Exam Tip
Do not limit AI to support. In scenarios, identify which value chain activity AI is helping and whether it supports continual improvement at any level.
AI and the Four Dimensions of Service Management
Dimension 1: Organizations and People
AI introduces new roles and skills. Culture must balance trust in AI with healthy skepticism to avoid automation bias and maintain accountability.
Dimension 2: Information and Technology
AI depends on quality data, integrated tools, and managed risks such as privacy, model drift, and security of AI components in the toolchain.
Dimension 3: Partners and Suppliers
Many AI services come from external providers. Contracts must clarify data use, transparency, and how AI incidents and updates are managed.
Dimension 4: Value Streams and Processes
AI can add or change process steps, such as automated triage. Value stream maps now include AI decision points and data flows explicitly.
Optimize and Automate: Applying the Guiding Principle to AI
Optimize Before You Automate
The principle says: fix and streamline the process first, then automate. AI on top of a broken process just speeds up bad outcomes.
Human Oversight
Use AI to recommend and support, not always to decide. For high-risk changes, AI can flag risks while humans remain accountable for approvals.
Good Automation Candidates
Automate frequent, low-risk, well-understood tasks like password resets or standard alerts. AI boosts speed and consistency here.
Measure and Improve
Track metrics like resolution time and satisfaction. Apply the continual improvement model to refine AI rules and models iteratively.
Worked Example: AI in Incident Management and AIOps
Retail Company AI Setup
A cloud retailer uses an NLP virtual agent, ML-based incident routing, and AIOps anomaly detection with automated remediation for simple infrastructure issues.
Impact on Utility and Warranty
Utility: 24/7 self-service and faster resolutions. Warranty: better SLA achievement by reducing response times and catching incidents earlier.
Value Chain Activities
Engage: virtual agent triages. Deliver & support: AIOps auto-fixes low-risk issues. Improve: teams analyze AI data monthly to refine knowledge and rules.
Four Dimensions View
People focus on complex work; data quality becomes critical; SaaS AIOps vendor is a key partner; incident value stream adds AI triage and automation steps.
Responsible AI in ITIL: Transparency, Fairness, Accountability
Why Responsible AI Matters
AI affects value, risk, and trust. In 2026, organizations align AI use with responsible AI principles, often influenced by regulations like the EU AI Act.
Transparency
Stakeholders should know when AI is used and how. In ITSM, clearly label AI-assisted routing or AI-generated summaries and document this in procedures.
Fairness
Avoid systemic bias, such as always deprioritizing certain users. Check training data and monitor outcomes to ensure equitable treatment in automated decisions.
Accountability
Humans remain accountable. ITIL practices must define ownership of AI decisions, error handling, and how AI-related risks are identified and controlled.
Thought Exercise: Spot the Responsible AI Issues
Work through this scenario mentally and note your answers (or jot them down) before revealing the explanations in your head.
Scenario: A university IT department introduces an AI system to prioritize incidents. The model is trained on the last 3 years of tickets. After deployment, staff notice:
- Tickets from the Business School are almost always marked as “high priority”.
- Tickets from student dorms are often marked “low priority”, even when they involve security issues.
- The portal does not inform users that AI is involved in prioritization.
- Analysts feel they cannot override priorities because their performance metrics reward closing “high-priority” tickets first.
Questions to think through:
- Which responsible AI aspects are being violated or stressed (transparency, fairness, accountability)?
- Which ITIL dimensions are impacted? Consider organizations and people, information and technology, partners and suppliers, value streams and processes.
- How could the continual improvement model be applied to fix this? Think step by step: What is the vision? Where are we now? Where do we want to be? How do we get there? Take action. Did we get there?
- Which guiding principles should guide the improvement? (Hint: think “focus on value”, “progress iteratively with feedback”, and “optimize and automate”.)
Pause and answer these before moving on to the next steps in the course later. This kind of analysis is exactly what exam case questions are testing, even if they do not mention “responsible AI” by name.
Quiz 1: AI, Value, and the Service Value Chain
Answer this question to check your understanding of how AI fits into the SVS and value chain.
A service provider introduces an AI-powered virtual agent that handles 40% of incoming support requests without human intervention. From an ITIL 4 perspective, which statement best describes this change?
- It only affects the Deliver & support activity because it automates ticket resolution.
- It supports multiple value chain activities, including Engage and Deliver & support, and must be evaluated in terms of its contribution to value co-creation.
- It creates a new standalone process in the service value system dedicated to AI operations.
- It replaces the need for continual improvement because the AI will keep learning automatically.
Show Answer
Answer: B) It supports multiple value chain activities, including Engage and Deliver & support, and must be evaluated in terms of its contribution to value co-creation.
The AI virtual agent clearly supports Engage (initial interaction with users) and Deliver & support (resolving requests). Its impact should be assessed in terms of value co-creation, such as improved outcomes and satisfaction. AI does not create a new standalone process in the SVS, and it does not remove the need for continual improvement; human-driven evaluation and refinement are still required.
Quiz 2: Optimize and Automate with AI
Check your understanding of the optimize and automate principle applied to AI.
Which approach best reflects the "optimize and automate" guiding principle when introducing AI-based incident categorization?
- Deploy the AI model immediately and let it learn from whatever data is available, then redesign the process later if needed.
- First standardize incident categories and update procedures, then train and deploy the AI model, monitoring its performance and adjusting over time.
- Rely entirely on AI for categorization and remove human review to avoid bias and inconsistency.
- Avoid using AI until the incident process is perfectly optimized and will never change again.
Show Answer
Answer: B) First standardize incident categories and update procedures, then train and deploy the AI model, monitoring its performance and adjusting over time.
Optimize and automate means improving and simplifying the process first (e.g., standardizing categories and procedures), then introducing automation such as AI. Performance should be monitored and refined through continual improvement. Fully removing human review can increase risk, and waiting for a "perfect" process is unrealistic and contrary to progress iteratively with feedback.
Key Term and Concept Review
Use these flashcards to reinforce essential ITIL and AI concepts in this module.
- service
- A means of enabling value co-creation by facilitating outcomes that customers want to achieve, without the customer having to manage specific costs and risks.
- service management
- A set of specialized organizational capabilities for enabling value for customers in the form of services.
- service value system (SVS)
- A model representing how all the components and activities of an organization work together as a system to enable value creation.
- service value chain
- A set of interconnected activities that an organization performs to deliver a valuable product or service to its consumers and to facilitate value realization.
- continual improvement
- A recurring activity performed at all levels to ensure that an organization’s performance continually meets stakeholders’ expectations.
- utility
- The functionality offered by a product or service to meet a particular need.
- warranty
- Assurance that a product or service will meet agreed requirements.
- How does AI typically support the Engage activity of the value chain?
- Through tools like NLP-powered virtual agents and chatbots that interact with users, capture intent, and route or resolve requests.
- Which guiding principle is most directly linked to introducing AI-based automation?
- Optimize and automate.
- Name two key aspects of responsible AI relevant to ITIL practices.
- Transparency (clear about AI use and decisions) and accountability (humans remain responsible for outcomes). Fairness is another important aspect.
Bringing It Together: Evaluating AI Use Against ITIL Principles
Start with Value
Ask: which customer or user outcomes does this AI support? Does it improve utility, warranty, or both? If value is unclear, the design is weak.
Principles and Iteration
Check optimize and automate and progress iteratively with feedback: was the process improved first, and are AI rules and models refined over time?
Holistic and Simple
Consider all four dimensions and avoid needless complexity. Good AI solutions fit people, data, partners, and value streams in a simple, practical way.
Responsible and Governed
Evaluate transparency, fairness, and accountability. Well-governed AI manages risks and supports trust, making it consistent with ITIL’s SVS.
Key Terms
- user
- A person who uses services.
- AIOps
- The application of artificial intelligence and machine learning techniques to IT operations data to detect patterns, predict issues, and support automated or assisted remediation.
- service
- A means of enabling value co-creation by facilitating outcomes that customers want to achieve, without the customer having to manage specific costs and risks.
- sponsor
- A person who authorizes budget for service consumption.
- utility
- The functionality offered by a product or service to meet a particular need.
- customer
- A person who defines the requirements for a service and takes responsibility for the outcomes of service consumption.
- warranty
- Assurance that a product or service will meet agreed requirements.
- generative AI
- AI models that can create new content (such as text, code, or images) by learning patterns from large datasets.
- service offering
- A description of one or more services, designed to address the needs of a target consumer group.
- value co-creation
- The joint activities performed by a service provider and a service consumer to create value.
- service management
- A set of specialized organizational capabilities for enabling value for customers in the form of services.
- service value chain
- A set of interconnected activities that an organization performs to deliver a valuable product or service to its consumers and to facilitate value realization.
- predictive analytics
- Techniques, often based on machine learning, that analyze current and historical data to make predictions about future events or behaviors.
- service value system
- A model representing how all the components and activities of an organization work together as a system to enable value creation.
- continual improvement
- A recurring activity performed at all levels to ensure that an organization’s performance continually meets stakeholders’ expectations.
- natural language processing (NLP)
- A field of AI focused on enabling computers to understand, interpret, and generate human language.