Organisations often favour technical solutions because they offer something tangible to implement. But complex transformation challenges involve interconnected decisions about how the business operates, how people work and what needs to change. Technology alone cannot resolve those questions.
Especially with artificial intelligence (AI), transformation can also affect the identity of organisational functions: their purpose, the expertise that defines them and the value they contribute. This is a critical consideration, not simply a side effect of technology adoption.
Many myths emerge because we confuse complex questions with complicated ones and initially focus only on technology adoption. That encourages the belief that implementing a tool will deliver the transformation.
Myth 1: AI improves the entire workflow
One common assumption is that artificial intelligence (AI) will improve the entire workflow simply because it improves individual tasks. A faster task does not necessarily mean a better overall process, particularly if it creates additional review, correction or coordination elsewhere.
That is why I would look at AI within the organisation’s operating model: its end-to-end processes, workflows and handovers between teams. This also means examining roles, responsibilities, decision-making and accountability for the overall outcome.
But I would go further. Technology creates opportunities to revisit both business and operating models, rather than simply fitting new tools into existing structures. We should ask not only how work could be done differently, but whether AI opens up different ways to serve customers and create value.
The question is therefore not simply “Where can we introduce AI?” but “What could we rethink or make possible that was not feasible before?”
Myth 2: A successful pilot predicts a successful rollout
The biggest reason is assuming that the pilot participants and conditions represent the wider organisation. A pilot involving AI enthusiasts in a controlled environment can produce results that do not translate directly into broader adoption.
I would treat pilot success as evidence of potential, not proof of readiness to scale. Before a wider rollout, I would test with people who have different levels of confidence, different responsibilities and different working conditions. The question is not only whether the solution works, but what support people need to make it work in everyday practice.
Myth 3: Agentic AI is a technology decision
Leaders first need to decide what agents may do autonomously and who remains accountable for their actions. That makes agentic AI a business and governance decision before it becomes a technology choice.
My starting questions would be: What outcome are we trying to achieve? Which decisions can we delegate, and which require human approval? What should happen when an agent makes a mistake or encounters something outside its remit? Choosing the technology should follow those decisions, rather than allowing the technology to define the boundaries of responsibility.
Myth 4: Resistance is a communication problem
Leaders should avoid assuming that hesitation means people do not understand or support the change. What looks like resistance may be a capacity problem, and better communication cannot compensate for change fatigue.
I would start by asking what else people are being expected to deliver, what competing changes they face and whether anything has been deprioritised. A useful question is: “What would need to change for this to become workable?” That opens a conversation about practical barriers rather than labelling people as resistant. The response may require clearer priorities, more support or different sequencing, not another communication campaign.
Myth 5: The dashboard knows
Yes. A dashboard can accurately show what happened while still leaving us with an incomplete understanding of why it happened. Metrics need interpretation, especially when reported results and people’s experience differ.
For example, I would not treat high usage alone as proof of value. I would also ask whether the work became easier, whether quality improved and whether effort was simply moved elsewhere. Conversations with employees and observation of everyday work can help explain the numbers. A dashboard should be the starting point for questions, not the end of the discussion.
Three questions every leader should ask
For leaders, moving beyond these myths starts with recognising that transformation is not simply a technology implementation, but a business challenge that requires their active involvement. With that in mind, I would ask leaders three questions: What future do we want to create for our organisation and customers? What new opportunities could technology unlock if we rethink our operating models? What needs to evolve across people, processes and technology to turn that vision into reality? The challenge is not answering them once, but revisiting them throughout the journey.
