Two conversations, a few weeks apart, told me almost the same story.
The first was with the General Manager of a Dubai-based property development company. The second was with the founder of a small BPO firm, also in Dubai. Different industries, different company sizes, no connection to each other. When I asked each of them, separately, what their AI strategy looked like, I got back nearly the identical scene.
Both described the same meeting. The CFO, sometimes a board member, asks: "So what's our AI strategy?" There is a brief pause. Someone mentions the ChatGPT Enterprise licences that finance signed off on. Someone else mentions the pilot marketing ran with a chatbot. Whoever is wearing the CIO hat that week says something about "exploring use cases." Everyone nods. The meeting moves on to the next agenda item.
Neither of them was uninformed. Neither was short on effort. And yet the honest answer, the one nobody said out loud in either room, was closer to: "We don't really have a strategy. We have activity."
I have now heard a version of this exact scene enough times, across industries as different as property development, business process outsourcing, banking, and mining, that I no longer think of it as an anecdote. It is a pattern. It is not a failure of intelligence or effort. It is what happens when a genuinely hard question gets asked in a room where admitting "I don't know" feels more dangerous than nodding along.
Why the nodding happens
Nobody wants to be the one who says "we don't have a strategy" in front of the board. So the room does what rooms do: it reaches for evidence of activity and presents it as evidence of direction. Licences purchased. A pilot running somewhere. A slide with the word "AI" on it from last quarter's town hall.
Activity feels like progress. It photographs well in a board pack. But activity and strategy are different things, and the gap between them is exactly where AI budgets quietly disappear, initiatives stall without anyone deciding to stop them, and, eighteen months later, the same question gets asked again in almost the same words.
Activity is what you can point to. Strategy is what you can be held accountable for. Most organisations have plenty of the first and very little of the second.
The three questions that actually separate strategy from theatre
You do not need a hundred-page AI strategy document to have a real answer. You need honest answers to three questions. In my experience, almost no leadership team can answer all three cleanly on the first attempt, and that is precisely the point: it is diagnostic, not a test you are expected to pass.
1. What business outcome are we actually trying to move?
Not "we want to use AI more." A specific, measurable result: reduce time-to-decision in underwriting by a third, cut manual data entry in finance operations by half, shorten customer response time from days to hours. If the honest answer is "we're not entirely sure yet," that is fine, it is also the first thing to fix, before another licence gets purchased or another pilot gets greenlit.
2. Who is accountable if this doesn't work?
Not a committee. Not "IT and the business together." One name. In most of the rooms I sit in, this question produces the longest silence of the meeting, because the honest answer is that accountability is genuinely diffuse. Everyone has a stake. Nobody owns the outcome. That diffusion is not a personality problem, it is a structural one, and it is entirely fixable once someone decides to fix it.
3. What happens when this goes wrong?
Every AI initiative eventually produces an error, an edge case, a decision someone questions. The organisations that handle this well decided, in advance, who gets told, what the escalation path is, and what "wrong" even means for this particular system. The organisations that handle it badly find out the process in real time, under pressure, usually in front of the board.
What changes when you can actually answer them
Nothing about this requires slowing down. In fact, the opposite is true. Leadership teams that can answer these three questions clearly move faster afterward, not slower, because they stop relitigating the same uncertainty in every subsequent meeting. The ambiguity that felt like flexibility was actually friction, and removing it is what lets initiatives that deserve momentum finally get it.
The next time someone in your leadership team asks "what's our AI strategy," you do not need a polished answer. You need the courage to treat it as a real question rather than a rhetorical one, and thirty honest minutes to work through the three questions above. Most organisations have never spent those thirty minutes. That is usually the entire gap.
Want an outside view on your own answers to these three questions? See our AI Strategy Advisory service.