Masterclass – AI Governance

Ask a senior leader how to govern AI, and the conversation rarely stays on just AI for long. It soon turns to a wider discussion about governance itself, and through painful memories of past challenges, to whether the organisation even knows how well it governs its technology decisions today.

The rapidly growing interest in AI across organisations of all sizes makes that question very hard to avoid. Various AI-related capabilities are introduced through free tools and browser sign-ups that no procurement process ever sees. These stealth-embedded AI capabilities can easily break the common assumption (and often, a requirement) that every decision can be traced back to a person who can explain it. Governance-as-designed and governance-as-practised may have drifted apart in the organisation long before AI arrived. The gap between them is where AI can amplify the unmanaged risks.

This masterclass is a practical working session focused on understanding and dealing with that gap. You will learn to assess how your organisation governs technology in practice, and where the formal structure and the day-to-day practice come apart, then stress-test both against the ways AI is already being used. The approach draws on the governance patterns model from ITIL (Version 5) and the AI-specific assessment tools described in the ITIL AI Governance white paper, worked through examples you can take back to your own organisation.

Facilitated by Kaimar Karu, this masterclass builds on the themes explored in the blog post “AI governance does not start with AI” and the itSMF UK webinar that followed, though it stands on its own and assumes no prior reading. Kaimar is a strategic adviser to senior IT leaders on governance, AI, and large-scale digital transformation. He served as Estonia’s Minister of IT and Foreign Trade, and earlier led the work at AXELOS (now PeopleCert) to reshape ITIL around modern ways of working.

He was one of the authors who worked on ITIL (Version 5), contributing to and reviewing content across the new version. He is the lead author of its Transformation publication, a co-author of Strategy, the creator of the governance patterns model used throughout this masterclass, and a co-author of the ITIL AI Governance white paper. He speaks internationally on governance, digital transformation, Cynefin, and AI ethics.

Target Audience

The masterclass is for anyone who carries responsibility for how technology and AI decisions get made and overseen, at any level. It deliberately spans a range of roles:

• CIOs, CDOs, and other senior IT leaders setting direction and accountable for the risks
• Heads of IT governance, risk and compliance leads who set the policies and procedures
• Service Owners who have to make governance work in practice
• Enterprise Architects, IT Service Managers, and hands-on practitioners who see, first-hand, where designed governance and real practice part ways

You need no deep technical AI knowledge. What helps most is some familiarity with how your organisation makes and oversees its technology decisions today – or at least thinks it does.

BOOK YOUR PLACE

What to expect

By the end of the masterclass, you will be able to:

• Assess which of the four governance patterns your organisation follows in practice, using the ten governance characteristics, and see where that differs from the assumed patterns
• Find the gap between governance-as-designed and governance-as-practised, and understand why AI widens it
• Use the 6C model to work out which AI capabilities carry which governance risks, and why e.g. a coordination-focused AI solution that triggers actions is a vastly different challenge from a clarification-focused AI solution that summarises documents
• Stress-test a governance pattern against a real AI use case and read where it breaks, whether on speed, scale, complexity, or uncertainty
• Walk through the four AI impact perspectives, with the risks and countermeasures for each
• Take away a repeatable method and a set of tools you can apply to your own organisation, and return to as your AI use grows

Because delegates come from different organisations, the exercises use shared, worked cases, so no one needs to bring their own confidential detail into the room. You practise the method during the day and leave able to run it on your own governance afterwards.

What will it involve?

The day moves from diagnosis through to what to change:

• The governance landscape (as per ITIL (Version 5)). The two axes of authority and assurance, and the four patterns they produce: Directive, Guided, Federated, and Autonomous
• Understanding your organisation’s patterns. The ten governance characteristics, how to plot them, and what clusters, outliers, and hybrid positions tell you
• The gap between formal and actual governance. Why what looks Directive on paper often runs as uncoordinated Federated or unsanctioned Autonomous in practice, and why that gap hides risk
• What AI changes. The two assumptions existing governance leaned on: a) that you could see the decision and b) trace it to an accountable person, and how AI breaks both
• The 6C model. Six functional AI capabilities (Creation, Curation, Clarification, Cognition, Communication, Coordination) and the different governance risks each one carries
• The four AI impact perspectives. Decision authority and risk management, ethical principles and responsible AI, data governance and performance management, and regulatory compliance and operational standards, each with its typical risks and countermeasures
• Stress-testing governance against AI. A worked example testing a governance pattern against a demanding AI use case, and how to read the breaking points
• The assessment approach. The four steps: stress-test existing governance, determine AI governance requirements, design and implement adjustments, and keep governance responsive as AI systems learn and change
• Adapting each pattern for AI. What Directive, Guided, Federated, and Autonomous governance each need to change to handle AI while keeping their character
• From oversight to stewardship. The maturity indicators, from ad hoc AI use to governance that adapts with experience
• Turning the assessment into a leadership conversation. How to take the pattern assessment and gap analysis into the next discussion about AI, and share the problem so it gets owned

Delivery Platform

This event will be delivered using Microsoft Teams.

PeopleCert ITIL CPD Points

itSMF UK members will automatically earn 7 ‘Professional Education’ ITIL CPD points for attending this event – more details available here.

How much does it cost?

Places are available to itSMF UK members for £295.00 + VAT.

Not yet a member? You’ll be shown a price of £460 + VAT which includes an ‘Individual’ level annual membership.  To find out about other membership levels, benefits and costs please click here.

REGISTER NOW!

Please contact the office on +44 (0) 118 918 6500 to discuss a dedicated delivery of this event for your organisation.
London