WHAT REALLY DRIVES AI ADOPTION? BEYOND LEADERSHIP AND ORGANISATIONAL CULTURE
This article explores how leadership, culture and organisational readiness interact, and why an organisation's capacity ultimately determines whether AI creates meaningful change.
A few days ago, a colleague read the case study I wrote on Coinbase (if you haven't seen it yet, I'll leave it here), and we ended up talking about it over a coffee. What he kept coming back to was the way its CEO thought, how he looked at risk, how that very particular way of standing in front of uncertainty seemed to filter into the way Coinbase brought in artificial intelligence. And as we talked it over, the question that started all of this came to me: to what extent does a company adopt AI because of how the person leading it thinks?
We finished the coffee without settling it, but the question kept turning over in my mind, and I carried it back to my desk.
The question, spelled out in full, was this: when a company adopts artificial intelligence, is it driven by the leadership at the top, or by the culture of the organisation? And behind that: are they two different things, or do they turn out to be the same thing seen from different angles?
I came in with a fairly firm hunch. It made sense to me that the weight would lie with leadership, that the way of thinking at the top would set the course and the rest would follow. So, I wanted to put that intuition to the test and see how well it held up against what we know about the subject today.
And it's worth looking at closely, because this goes beyond curiosity: if you misunderstand how adoption happens, you're very likely to end up designing your strategy for it badly too.
So, I set about pulling together what's known on the subject and cross-checking it against what I see in practice, to see how well my intuition held up.
if you misunderstand how adoption happens, you're very likely to end up designing your strategy for it badly too.
The first thing I found was clear: there is no single factor that explains adoption. Neither the CEO, nor the culture, nor competitive pressure works on its own as the answer. Every time a company brings in AI, there are several things acting at once: leadership, culture, the resources it has, its size, the infrastructure already in place, suppliers, the environment it operates in, and how ready it is to take the step.
So, the question began to change shape. It stopped being "which factor explains adoption?" and became "how do all these factors combine?".
Now, the fact that there are many factors doesn't mean they all carry the same weight. And there's one that comes up more than the rest: how prepared the organisation is to bring in the technology.
And this surprised me, because I expected to find a story about people, a visionary leader, a team with the right mindset, but I found a story about conditions. And readiness means having the infrastructure, the resources and the people with the capabilities to bring in AI, among other factors.
I expected to find a story about people. I found a story about conditions.
With that on the table, I went back to my original question: where, then, does leadership sit?, and here the picture got interesting. In some studies leadership shows up as one of the best predictors of adoption; in others it comes in fourth, considerably further down; and in several, the organisation's readiness comfortably outweighs it. The same question, different answers depending on where you look.
When people talk about leadership, they're not always talking about the same thing. For some, leadership is the backing of top management, the leadership supporting it and putting up the resources. For others, it's transformational leadership, the kind that inspires and gives vision. And for others, it's digital leadership: steering the meeting point between the technological and the human in the company's transformation, a term that brings together many ideas and is still taking shape.
Three different things travelling under a single word, so it makes complete sense that they arrive at different conclusions. And the same happens with culture: each person measures a different version, and then we compare results as if they were talking about the same thing.
In the end I settled on a measured and honest conclusion: leadership matters, everyone confirms that. What changed was the magnitude: it went from seeming like the main factor to being one among several.
On top of that, we almost always assume that companies adopt AI under pressure from the competition or the market, whether because the one next door is moving ahead, or because of imposed regulation. It sounds so logical on the face of it; and yet, when it's measured on AI's own ground, that competitive pressure shows up among the weakest factors. In several studies it barely moves the needle, and in some it doesn't even weigh enough to count.
I thought about it for a while, and I want to offer a reading of my own, not something proven: perhaps market pressures don't disappear, but rather come in through another door. They reach the people leading the company, and that's where it gets decided what to do with them. The pressure still carries weight, only it doesn't do so directly: it passes first through leadership, through how those at the top read that context and choose to respond. And here I'll leave something to think about as a hypothesis: those outside pressures (the competition, the market, the rules of the game) may well go on shaping the kind of leadership that gets exercised; that leadership, in turn, shaping the culture; and that culture ending up tilting adoption. Almost without anyone noticing.
Market pressures still matter. They just don't act directly; they pass first through leadership.
And piecing it together, a reading of my own took shape. The studies give them different names; readiness, resources, infrastructure, capabilities, but to me they all began to sound like the same underlying thing. I call it capacity: what an organisation is able to do with the technology it brings in. And if that's right, then leadership and culture stop competing with each other and become two different ways of building that same capacity, one from the person who leads, the other from the organisation's way of being.
With that thread in hand, something appeared that made sense of everything else. Plenty of companies buy AI, launch pilots, train their people, pay for licences, kick off projects, and even so the organisation goes on running more or less as it did before. And we've seen this before: it happened with the big management systems, with the cloud, with big data, with Industry 4.0.
The technology comes in fast, but the results arrive later and at their own pace. Some have given that shape a name: the J-curve. At first productivity actually dips, because the company is rearranging itself, redesigning processes, adding capabilities, adjusting the way it works and only later does it climb, once those deeper changes are in place. First, it falls, then it takes off, and it's exactly the distinction I've been writing about: measuring the use of AI is one thing, how many use it, how many pilots, how many training sessions, all easy to measure, and measuring its impact, what actually changed on the inside, is quite another.
Across the successive waves of digital transformation, the discussion was never really "is it the CEO or is it the culture?". What comes into view is a fuller reality: many factors combine to explain adoption, and adoption on its own has never guaranteed impact.
If someone asked me today, over another coffee, whether AI adoption is mainly about the CEO or the culture, I'd probably answer differently. I think both matter, but I no longer think that's the most useful question. The deeper question is whether the organisation has built the capacity to turn AI adoption into organisational change. That's the thread I kept finding, even when it appeared under different names: readiness, capabilities, infrastructure, resources. To me, they all describe different ways of talking about the same underlying capacity. Leadership and culture are not competing for that place, they're two ways of building it.
If you'd like to apply these ideas to your own organisation, I've put together two short assessments. One is designed for organisations preparing to adopt AI, the other is for organisations that have already adopted AI and want to evaluate its impact. Both are free, take less than five minutes, and include a personalised summary with one practical next step.
If this article sparked a question or a different way of thinking, I'd love to hear it. You can write to me anonymously below. I read every submission, and many of them would inspire future articles.