I have spent much of my career leading organisational transformation and helping businesses navigate complex change. One lesson has remained remarkably consistent: transformation rarely fails because organisations lack a strategy or invest in the wrong technology. It fails when they underestimate what it takes to change how people work, how decisions are made, the culture and how organisations operate.
That challenge has become even more pronounced with the arrival of AI. Unlike previous waves of technological change, AI is evolving at a pace that makes traditional approaches to transformation hard to sustain. Organisations can no longer treat change as an occasional project with a defined beginning and end. They need to develop the ability to continuously adapt, learn and redesign how they operate, in real time.
Drawing on these experiences, I want to share five principles that I believe are essential for organisations to navigate change successfully in the age of AI, and that we at Bridge Advisory help organisations put into practice.
The Failure of Traditional Transformation
For decades, organisations treated transformation as an episodic event: a company modernised its systems, trained its workforce, and operated in a stable environment until the next wave of disruption arrived. But today, AI is advancing at a velocity that far outpaces traditional transformation. Five-year strategic roadmaps, annual planning cycles and rigid top-down management have become major liabilities.
Investment is not the problem, particularly when it comes to AI. However, according to Boston Consulting Group, only 26% of companies have developed the capabilities to move beyond proofs of concept and generate tangible value from AI.
The primary bottleneck is not technical infrastructure or algorithms: BCG found that around 70% of the challenges organisations face in implementing AI stem from people and process issues. Conventional top-down campaigns rarely close this gap because leaders are often too far removed from daily operations to know how work will actually change, and employees respond with hesitation rather than commitment. To thrive, organisations must move beyond episodic change management and build a lasting capability for change resilience.
1. Building Change Resilience: The Three Core Muscles
Change resilience is the ability to turn continuous disruption into repeatable learning and improvement. Rather than treating change as a project with a fixed beginning and end, resilient organisations continuously strengthen three capabilities:
Diagnose: The ability to identify weak technological signals, unoptimised processes and competitive and societal signals early, before disruption forces a crisis.
Redesign: The capacity to reshape workflows, roles and decision rights, and to redeploy capital, data and talent behind them, in days or weeks rather than waiting for annual budget cycles.
Confirm: The discipline to codify what works (into playbooks, operating procedures and policies) so that each new initiative starts from a higher baseline rather than reinventing the wheel.
These capabilities are not about pushing AI as far and as fast as possible. Each organisation should progress at the pace and level of AI autonomy that suits its value, complexity and risk, while continuing to evolve its operating model as AI capability grows.
2. Moving Beyond Pilots: The "Deep and Narrow" Strategy
A common failure mode in AI-driven transformation is spreading resources thinly across multiple disconnected pilots. This "shallow and broad" approach yields minor, localised time savings but fails to deliver strategic advantage, and it often triggers change fatigue among staff.
Instead, high-performing organisations adopt a "deep and narrow" approach. They select a single critical domain, such as end-to-end procurement or marketing operations, and fundamentally reimagine every interlinked workflow. While recent research suggests that only 4% of enterprises currently execute change this way, those that do achieve twice the ROI because deep transformation builds protective, hard-to-copy capabilities embedded in how the organisation works.
3. Integrating Digital Labour into Hybrid Teams
Redefining change management requires expanding the definition of the workforce itself. Emerging agentic AI models are evolving from simple assistants into "digital teammates" capable of carrying out complex, multi-step tasks.
Leaders must deconstruct traditional job roles into individual tasks and decide, workflow by workflow, what AI owns, what people own and what is shared. High-volume data validation may sit with digital labour; other tasks require human judgement. Clear decision rights, guidelines and escalation rules between AI agents and human specialists prevent operational conflict and make it obvious who acts when something goes wrong.
Crucially, people remain accountable for outcomes. As AI takes on more of the work, a significant focus must be on how people stay in the loop: where they exercise judgement, where they supervise AI output, where they intervene, and where they own the critical decision points. Roles are not simply elevated towards "higher-value" work; they must be redesigned around these new responsibilities.
4. Frontline Ownership of Change
Change cannot be dictated from the corporate HQ; it must be co-created with the people closest to the daily work. Shifting from leader-led planning to employee-owned implementation increases the probability of achieving the project’s goals, boosts discretionary effort and significantly reduces execution time.
In practice, this means involving diverse frontline perspectives early. It also means letting teams set their own implementation goals and identify obstacles within clear strategic guidelines.
This involvement is itself the engagement. Rather than relying on a better communications campaign that sells benefits and downplays risks, two-way conversations about how work will change build capability, surface hidden concerns early, and give people a real sense of agency during uncertain transitions.
5. Leading with a Beginner's Mindset
As AI challenges traditional expertise, the instinct for leaders and professional experts is to protect their authority. However, by doing so they might erode flexibility and foster confirmation bias. To navigate rapid technological shifts, leaders must cultivate Shoshin (a Japanese concept from Zen Buddhism that translates to "beginner's mindset"), approaching novel tools with intellectual humility, openness and curiosity.
Research suggests that leaders who demonstrate openness, encourage experimentation and regularly check in with their teams create more positive working environments. Employees who experience this type of leadership report 25% less burnout and are significantly less likely to quit.
Conclusion:
Redefining change management in the AI era is not about implementing a technology upgrade; it is about building an organisation that can continuously renew itself.
By combining organisational resilience, redesigned human–AI workflows, frontline ownership and adaptive leadership, organisations can move beyond exhausting one-off transformations and develop the capacity to thrive in continuous disruption.