The UAE leads the world in AI adoption. Almost no organisation here has rebuilt anything

In the first quarter of this year the United Arab Emirates became the first economy in the world in which more than seven in ten working-age people use artificial intelligence. The figure is 70.1%. The global average is 17.8%, and large parts of the world remain below 10%.

It is a genuinely remarkable number and it was widely reported as such. It is also, read on its own, one of the more misleading statistics currently circulating in boardrooms here, because of what it does not measure.

What the number counts, and what it does not

It counts people. It does not count organisations, and it does not count work.

An individual using a language model for ninety minutes a month is counted, and there is nothing wrong with that as a measure of diffusion - it is precisely what the research set out to establish. But the thing a chief commercial officer needs to know is a different question entirely: how much of the work inside the organisation has actually been rebuilt around this, as opposed to being carried out in the same sequence as before by people who now have a faster way of performing one step of it.

On that question the answer here is much closer to everyone else's. Roughly 7% of organisations have integrated AI into their core operations in a way that changed how the work is done. The remainder have licences, enthusiasm, and a largely unchanged process map.

Why the two numbers diverge so sharply

Three reasons, and none of them is a failure of ambition.

Individual adoption requires nobody's permission. A marketing executive who begins using a language model to draft social copy has made a decision affecting one person's working hours. Redesigning how the campaign reporting process works affects a dozen people, a platform, an agency contract and a monthly meeting. The first happens in an afternoon by itself; the second requires somebody senior to decide it should happen and to carry the disruption while it does.

The usage is invisible to the organisation. A great deal of the AI use in commercial teams in this region is informal and unreported, conducted on personal accounts because the corporate position was unclear and nobody wanted to ask. That has two consequences: the organisation cannot see where the value is being created, so it cannot invest behind it, and the data governance question that should have been answered at the outset has instead been answered accidentally by two hundred individual decisions.

The spending went to licences rather than to redesign. Enterprise AI investment in the UAE rose by 105% in a single year, and the overwhelming majority of it bought tools. Very little of it bought the work of changing what the tools are used for. This is the most expensive version of the mistake, because the licence cost recurs annually whether or not anything downstream of it changed.

What this looks like from inside a commercial function

Concretely, and this pattern repeats with remarkable consistency across the organisations I speak to in Dubai:

• -Most of the marketing team uses AI daily. The monthly report is still compiled by hand, by the same person, over the same two days.
• -Sales people draft emails with it. The proposal process still takes three days, because the delay was never in the drafting.
• -The agency retainer has not been renegotiated, although the production half of its scope has become considerably cheaper to deliver.
• -Nobody has written down what may and may not be placed into a public model, so everybody has decided individually.
• -The organisation holds more licences than it has people actively using them, and cannot say which of the two figures is the problem.

Every item on that list is a redesign question rather than a technology question, and every one of them is invisible in a statistic that counts individual usage.

Why this is an advantage rather than a criticism

It would be easy to read the above as an argument that the region is behind. It is the opposite.

A workforce that already uses these tools daily is the expensive precondition, and it is the one most markets do not have. Organisations in Europe attempting this work spend a substantial portion of the effort on persuasion - on convincing people to try something they are suspicious of. That problem does not exist here. The people are ready; the processes around them are not. That is a much better position to be in, because redesigning a process is a decision somebody can take on a Tuesday, whereas changing a workforce's disposition towards a technology is a multi-year cultural project.

It also means the advantage remains available, which will not be true indefinitely. When 7% of organisations have restructured and the other 93% have not, the gap in cost per output compounds every quarter in favour of the few. The moment the proportion inverts, doing this work stops being an advantage and becomes a condition of remaining competitive, and it is considerably less pleasant to undertake under those circumstances.

What I would do about it

Stop measuring licences and start measuring work. The question at the next commercial review should not be how many people have access, which is a procurement metric, but which specific workflows are performed differently than they were a year ago, and what the difference is in hours and in cost per output.

In most organisations the honest answer to that question is none, or one. That is not a reason for discomfort - it is the same answer that 93% of organisations would give. It is, however, a considerably more useful starting point than a headline adoption figure that describes the country rather than the company.