AI
AI Transformation Through Kotter’s 8-Step Change Model
Introduction
Artificial intelligence is rapidly moving from experimentation to operational deployment. Yet despite the excitement, many AI initiatives struggle to deliver lasting impact. Industry analyses from firms such as KPMG suggest that only a minority of organisations successfully scale AI beyond pilot projects or proof-of-concepts.
The problem is rarely the technology. Decades before AI entered the workplace, leadership scholar John P. Kotter showed that large-scale transformations fail not because of inadequate tools, but because of weak urgency, fragile coalitions, unclear vision, and declaring victory too early. His analysis, captured in Leading Change: Why Transformation Efforts Fail, has since been discussed and critiqued in professional and academic settings, including journal club reviews that highlight the risks of “premature victory” and underestimating cultural change.
More recently, researchers have begun contextualising Kotter’s 8-step model to sustainable digital transformation, mapping concrete activities and leadership behaviours onto each step in technology-driven change programmes. This work suggests that classic change frameworks still provide useful scaffolding, but must be interpreted in light of new technologies, stakeholder expectations, and sustainability pressures.
At the same time, scholarship on AI-driven organisational change argues that AI reshapes structures, workflows, decision-making, and culture, not just tools. AI systems are adaptive, their performance evolves with data and context, and they demand ongoing adjustments to skills, governance, and operating models.
Kotter’s eight-step model provides a structured approach to leading this kind of change. As summarised by The Open University, the steps are:
- Create a sense of urgency
- Build a guiding coalition
- Form a strategic vision and initiatives
- Enlist a volunteer army
- Enable action by removing barriers
- Generate short-term wins
- Sustain acceleration
- Institute change
Together with more recent work on digital and AI-driven organisational change, these steps describe how organisations move from recognising the need for change to embedding new ways of working into everyday operations.
Example Use Case
To make this concrete, this article reinterprets Kotter’s model for the AI era, drawing on research that contextualises the 8 steps for digital transformation and on emerging theory about AI-driven organisational change. It follows a fictional organisation, Northstar Manufacturing, as it moves through the eight steps of AI transformation.
1. Create a Sense of Urgency
What Kotter meant: Transformation begins when organisations recognise that maintaining the status quo is riskier than changing.
What this means for AI: Urgency comes from missed opportunities to augment human performance, not just fear of automation.
In AI transformation, urgency is rarely driven by fear alone. More often, it emerges from a growing gap between how work is currently done and what is now possible. This tends to show up in small but compounding inefficiencies—manual reporting, delayed decisions, and an inability to act on available data.
Over time, these gaps become visible not just as operational friction, but as missed opportunities to improve performance without increasing headcount.
Example Use Case
At Northstar, leadership recognised that competitors were using AI to optimise production schedules, reduce downtime, and accelerate customer response times, while internal teams relied on manual spreadsheets and siloed systems.
Urgency emerged from:
- rising expectations for real-time insights
- increasing data volumes overwhelming existing processes
- productivity demands without additional headcount
- customer frustration with slow turnaround times
Leaders made this tangible by highlighting specific inefficiencies: a weekly reporting process consuming 18 hours, delays in identifying machine failures, and fragmented data limiting decision-making.
The goal was clarity, not alarm. Employees needed to understand why change was necessary—and why it mattered now.
2. Building a Guiding Coalition
What Kotter meant: Change requires a group with enough influence and credibility to lead it.
What this means for AI: AI transformation must be cross-functional, not confined to IT.
AI initiatives often stall when they are treated as purely technical programmes. In practice, successful transformation requires coordination across operations, data, people, and governance.
A guiding coalition creates alignment across these domains, ensuring that decisions are not optimised locally but support a broader shift in how the organisation operates.
Example Use Case
Northstar initially positioned AI as an IT-led initiative. This quickly proved insufficient.
The coalition expanded to include:
- operations leaders who understood workflow bottlenecks
- data specialists assessing feasibility
- HR leaders focused on capability building
- frontline supervisors influencing daily behaviour
This group aligned priorities, removed obstacles, and modelled the behaviours required for transformation: experimentation, transparency, and adaptability.
3. Form a Strategic Vision and Initiatives
What Kotter meant: A clear vision aligns effort and guides decision-making.
What this means for AI: The goal is not isolated use cases, but a coherent operating model.
Many organisations approach AI through disconnected pilots. While these can demonstrate potential, they rarely add up to meaningful transformation without a unifying direction.
A strong AI vision defines how work should change: what gets automated, how decisions are made, and how teams interact with systems.
Example Use Case
Northstar had previously run disconnected AI pilots. None contributed to a broader transformation.
The coalition reframed the vision around augmentation:
- AI automates repetitive work
- teams gain faster access to insights
- employees can design and adapt workflows directly
From this, they prioritised:
- automated production reporting
- predictive maintenance dashboards
- tools enabling business teams to build internal applications
Each initiative was selected based on how clearly it moved the organisation toward the future state.
4. Enlist a Volunteer Army
What Kotter meant: Change accelerates when large numbers of people actively support it.
What this means for AI: Adoption grows through participation and experimentation, not mandates.
AI adoption is difficult to enforce top-down. Employees need to see how tools improve their own work before they commit to using them.
Participation creates momentum. As more individuals experiment, capability spreads organically across the organisation.
Example Use Case
At Northstar, early adopters emerged organically—engineers, analysts, and supervisors experimenting with AI tools.
To scale this:
- internal AI workshops were introduced
- cross-team experimentation was encouraged
- a community of practice was formed
- hackathons focused on real challenges were launched
As participation increased, curiosity evolved into capability. Adoption became peer-driven rather than top-down.
5. Enable Action by Removing Barriers
What Kotter meant: Structural obstacles must be removed to enable progress.
What this means for AI: The biggest barriers are often data, governance, and rigid processes—not algorithms.
Even with strong intent, teams cannot act if systems are fragmented, rules are unclear, or processes are too slow. Removing these barriers is what turns strategy into execution.
Example Use Case
Northstar identified key blockers:
- fragmented data across legacy systems
- uncertainty around responsible AI use
- slow, approval-heavy processes
The coalition responded with:
- a clear AI governance framework
- simplified data access policies
- guidelines for safe experimentation
- a centralised platform for building AI workflows
These changes enabled teams to act with confidence.
Operationalising AI Transformation
This is the inflection point between removing barriers and generating results.
Many organisations reach alignment on vision but struggle to execute because their systems cannot adapt. Traditional enterprise software often makes workflows rigid and slows experimentation.
AI transformation begins to scale when organisations treat AI not as a set of tools, but as an operational capability—one that allows teams to design, adapt, and improve how work gets done.
Example Use Case
Teams were enabled to:
- design and modify workflows directly
- embed AI into everyday processes
- connect data, automation, and decision-making
Engineers built dashboards, analysts automated reporting, and supervisors improved workflows without long development cycles.
6. Generate Short-Term Wins
What Kotter meant: Early successes build credibility and momentum.
What this means for AI: Wins should demonstrate tangible improvements to everyday work.
Short-term wins are critical in AI because benefits can otherwise feel abstract. Visible improvements help build trust and justify continued investment.
Example Use Case
Northstar focused on measurable outcomes:
- automating reporting saved 18 hours per week
- AI-assisted triage reduced response times by 30%
- predictive models improved failure detection
These wins were shared widely, reinforcing progress and encouraging further adoption.
7. Sustain Acceleration
What Kotter meant: Change must build continuously; early success is not the endpoint.
What this means for AI: Transformation is ongoing, not project-based.
Many organisations lose momentum after early wins, slipping back into familiar ways of working. Sustaining acceleration requires continuous investment, reinforcement, and adaptation—precisely the challenge Kotter described when he warned against declaring victory too early.
Example Use Case
Northstar expanded successful use cases into new departments, pairing each rollout with training, governance updates, and clear ownership.
The organisation also invested in ongoing model monitoring, retraining, and performance review, ensuring that AI systems remained effective over time.
8. Institute the Change
What Kotter meant: New approaches must be embedded into organisational culture.
What this means for AI: AI must become part of how work is done, not an optional tool.
Sustained transformation requires structural change. This includes updating roles, metrics, governance, and funding models so that AI is fully integrated into the organisation.
Example Use Case
Northstar embedded AI into everyday operations by:
- redesigning roles around oversight and exception handling
- updating KPIs to reflect automation and augmentation
- allocating budget for ongoing model maintenance
- establishing regular governance forums
These changes ensured that AI was no longer a project, but part of the operating model.
Why Change Frameworks Still Matter in the AI Era
AI introduces new capabilities, but the challenges of adoption—alignment, communication, trust, and behaviour—remain fundamentally human.
Research on sustainable digital transformation shows that Kotter’s eight steps can be contextualised to technology-intensive change, with step-by-step activities mapped to each phase.
At the same time, work on AI-driven organisational change highlights how AI reshapes structures, workflows, and decision-making, requiring ongoing adaptation rather than one-off projects.
Frameworks like Kotter’s continue to provide a useful structure for navigating transformation, even as organisations adapt them to the realities of AI.
Find out more About Cyferd
New York
Americas Tower
1177 6th Avenue
5th Floor
New York
NY 10036
London
2nd Floor,
Berkeley Square House,
Berkeley Square,
London W1J 6BD
