AI-G04 — AI Transformation and Change Management

Course Code: AI-G04  |  School: School of Artificial Intelligence  |  Cluster: Level 4 — Governance & Policy

Level: Advanced  |  Duration: 6 weeks · 24–30 learning hours  |  Language: English  |  Certificate: Executive Certificate (non-degree)  |  Format: Self-paced with AI support under human supervision

Overview

Most AI programmes fail for organisational reasons rather than technical ones. The model works, the pilot succeeds, and then nothing changes, because the people whose work would have to change were never given a reason, a skill, or a say. This course treats that as the central problem.

The material is drawn from the established literature on organisational change and diffusion, applied to a specific case: technology that alters how professional judgement is exercised. That specificity matters. Resistance to AI in a university or a ministry is frequently not fear of redundancy but a well-founded professional objection to being made accountable for output one cannot verify. A change programme that treats that objection as irrationality will fail, and deserves to.

Learners build a full change plan for one real adoption in their own institution, including stakeholder analysis, a communication plan, a capability programme, a redesign of the affected roles, and a measurement plan that would detect both adoption and quiet abandonment.

Learning outcomes

On completion, a successful learner will be able to:

1. Analyse stakeholders by interest, influence and the specific loss each faces, distinguishing legitimate professional objection from status protection.
2. Diagnose why a previous technology initiative in your institution succeeded or failed, using evidence rather than folklore.
3. Design a communication plan that states honestly what will change, including for whom the change is unwelcome.
4. Redesign affected roles so that accountability, verification duty and workload are coherent after adoption.
5. Build a capability programme sequenced to the adoption, with supervision competence as an explicit target.
6. Specify a measurement plan that detects genuine adoption, workaround behaviour and quiet abandonment.
7. Sustain the change: ownership, review, and the conditions under which the initiative should be reversed.

Who this course is for

Directors and deans leading adoption, heads of human resources and organisational development, programme managers, union and staff representatives engaged in consultation, and consultants running transformation work.

Prerequisites

AI-F01 and either AI-G01 or AI-G02, or equivalent institutional experience. Learners must have a real, current or imminent adoption to work on; hypothetical cases are not accepted for the assessed work.

Syllabus

Module 1 — Why the pilot succeeded and nothing changed

Focus. The gap between demonstration and institutionalisation. Studying the learner’s own institutional history of technology initiatives, and extracting the actual mechanism of failure.

Lessons. 1.1 Demonstration, pilot, adoption, institutionalisation. 1.2 Post-mortem of a local initiative. 1.3 Recurring mechanisms of failure. 1.4 What would have had to be true.

Core reading. John P. Kotter, Leading Change, revised edition (Boston: Harvard Business Review Press, 2012), chapters 1–3.

Deliverable. Post-mortem of one previous initiative in your institution, with the failure mechanism named.

Module 2 — Stakeholders and the anatomy of objection

Focus. Mapping interest, influence and loss. Separating professional objection grounded in accountability from resistance grounded in status or workload, because the two require different responses.

Lessons. 2.1 Interest, influence, loss. 2.2 Legitimate professional objection. 2.3 Status, workload and identity. 2.4 Consultation that is not theatre.

Core reading. Everett M. Rogers, Diffusion of Innovations, 5th edition (New York: Free Press, 2003), chapters 1 and 6–7.

Deliverable. Stakeholder analysis with a documented objection register.

Module 3 — Communication that survives scrutiny

Focus. Saying what will actually change. Sequencing announcements, handling the question of job security truthfully, and the cost of overpromising to secure early support.

Lessons. 3.1 What must be said first. 3.2 Answering the redundancy question honestly. 3.3 Overpromising and the credibility debt. 3.4 Channels, cadence and feedback loops.

Core reading. Kotter, Leading Change, chapters 4–6.

Deliverable. Communication plan with core messages and the answers to the five hardest questions you expect.

Module 4 — Role redesign and the verification duty

Focus. If a system produces output that a member of staff must sign, that verification is work and must appear in the role. Redesigning duties, accountability and workload so the arrangement is honest.

Lessons. 4.1 Verification as work, not attitude. 4.2 Accountability after automation. 4.3 Workload arithmetic. 4.4 Job design and consultation obligations.

Core reading. Thomas H. Davenport & Julia Kirby, Only Humans Need Apply: Winners and Losers in the Age of Smart Machines (New York: Harper Business, 2016), chapters 3–5.

Deliverable. Redesigned role descriptions for the two most affected positions, with workload accounted for.

Module 5 — Capability, sequencing and supervision competence

Focus. Training that arrives before it is needed rather than after, targeted at supervisory judgement rather than tool operation. Identifying and supporting internal champions without creating dependency.

Lessons. 5.1 Sequencing capability to adoption. 5.2 Supervision competence as the training target. 5.3 Champions, and the risk of single points of failure. 5.4 Support in the first ninety days.

Core reading. Rogers, Diffusion of Innovations, chapters 8–9. Tim Fountaine, Brian McCarthy & Tamim Saleh, “Building the AI-Powered Organization”, Harvard Business Review (July–August 2019).

Deliverable. Capability programme with schedule, targets and assessment method.

Module 6 — Measurement, sustainment and reversal

Focus. Indicators that distinguish adoption from compliance theatre, detection of workarounds and quiet abandonment, ownership after the programme closes, and the conditions under which the change should be reversed.

Lessons. 6.1 Adoption indicators versus activity indicators. 6.2 Detecting workarounds and abandonment. 6.3 Ownership and review after handover. 6.4 Reversal conditions and how to state them.

Core reading. Kotter, Leading Change, chapters 7–10. George Westerman, Didier Bonnet & Andrew McAfee, Leading Digital (Boston: Harvard Business Review Press, 2014), chapters 8–10.

Deliverable. Final submission: complete change plan including measurement, sustainment and stated reversal conditions.

Assessment

Pass mark 70 per cent. All assessed components must be attempted. Every mark in this course is issued by a human assessor; no assessment outcome is generated automatically.

Rubric criteria

Each assessed artefact is marked against four criteria at four levels (distinction, pass with merit, pass, fail).

1. Diagnostic honesty: does the post-mortem identify a mechanism that implicates decisions rather than only circumstances?
2. Treatment of objection: are legitimate professional concerns engaged on their merits rather than managed away?
3. Operational realism: does the plan account for workload, sequencing and capacity that actually exists?
4. Falsifiability: are adoption indicators and reversal conditions specified such that failure would be visible?

Reading list

Core. John P. Kotter, Leading Change, revised edition (Boston: Harvard Business Review Press, 2012). Everett M. Rogers, Diffusion of Innovations, 5th edition (New York: Free Press, 2003).

Applied. Thomas H. Davenport & Julia Kirby, Only Humans Need Apply (New York: Harper Business, 2016). George Westerman, Didier Bonnet & Andrew McAfee, Leading Digital (Boston: Harvard Business Review Press, 2014).

Articles. Tim Fountaine, Brian McCarthy & Tamim Saleh, “Building the AI-Powered Organization”, Harvard Business Review (July–August 2019). Thomas H. Davenport & Rajeev Ronanki, “Artificial Intelligence for the Real World”, Harvard Business Review (January–February 2018).

Working paper. Erik Brynjolfsson, Daniel Rock & Chad Syverson, “Artificial Intelligence and the Modern Productivity Paradox”, NBER Working Paper 24001 (2017).

All items are published works identifiable by author, title and publisher. Learners obtain them through an institutional library or the publisher. The Academy does not distribute copyrighted texts.

Academic integrity and use of AI

Generative tools may be used in producing assessed work under three conditions. Use must be disclosed in a short statement appended to each submission, naming the tool and the task it performed. Any factual or technical claim originating from a generative tool must be verified against a citable source before it enters assessed work, and the verification must be evidenced. The analytical judgement in each artefact must be the learner’s own and must be defensible in a short follow-up. The change plan must concern a real adoption in your own institution. Plans written for a fictional organisation cannot be assessed against the criteria above.

Instructor: pending owner confirmation. Pricing: pending owner approval. Reference list verified against publisher records; any later addition is marked for verification before publication.