The Adoption Chasm

84% bought it. 11% make money from it. The whole story is in the gap between those two numbers. Idle software does not compound. A parked tool is not lag waiting to resolve, it is a slide that says AI-enabled and earns nothing. Adoption is a starting gun the region keeps mistaking for a finish line.

The headline is that 84% of Gulf companies now use AI in at least one business function, up from 64% in 2023. Keynotes quote it. Board decks quote it. It is read as proof the region has arrived. It proves the opposite. In the same survey, only 31% of those firms have scaled AI beyond pilots. The group McKinsey calls value realizers, firms that have scaled and attribute at least 5% of earnings to AI, is just 11%. So 84% bought it, and 11% make money from it.

A bought tool is a line item. A built capability is a moat. The region is celebrating the line item.

The Conventional Reading Is Wrong

The comfortable interpretation goes like this. Adoption is the first step. Value is the second. The Gulf is early on a known curve, and time alone carries the 84% to the 11%. Patience. That reading is wrong, and expensively so. Time is not the missing ingredient.

Name the gap. I call the 73-point space between those two numbers the Adoption Chasm: the distance between buying a tool and building a capability. The chasm is not a delay. It is a different kind of work. Buying a tool is a procurement decision. It clears a budget line, closes in a quarter, and produces a logo on a slide. Building a capability is an operating-model decision. It requires three things no licence includes.

1. The judgment hire. Someone senior enough to decide which decisions the model is allowed to touch, and which it is not.

2. The process redesign. The workflow gets rebuilt around the model, not the model bolted onto a workflow designed for humans. A faster horse is still a horse.

3. The conviction to change the operating model. Reporting lines, incentives, and accountability move so the capability survives contact with the org chart.

You cannot purchase any of the three. You build them. That is why the chasm exists, and why it does not close on its own. The chain is tight. A tool produces output. A capability produces outcomes. Output is what the software does. Outcome is what the firm earns. The 73 points between 84% and 11% are precisely the firms that bought output and have not built the capacity to convert it into earnings. Adoption is not transformation, and McKinsey's own numbers prove it.

The Gulf Is Not a Beginner

Here the regional read gets interesting, and the just wait thesis quietly collapses. The Gulf is not uniformly stuck in the chasm. In places it is already fluent. It just is not fluent in the places that show up in an enterprise AI survey.

Consider this year's Hajj. Saudi Arabia closed the season having used AI-driven operations to orchestrate roughly 1.7 million pilgrims across five days and tens of square kilometres of dense, time-critical movement. Read that as an operations problem, not a press release. A country coordinating 1.7 million people in real time, against hard physical and religious deadlines, where failure is measured in human lives, is not running a pilot. You do not pilot at that scale. You do not pilot at that consequence. You operate at that scale only when the capability is already built into the institution.

That is fluency. It tells you the ceiling in the region is not the technology. The ceiling is whether the capability gets built deliberately or left to accrete by accident.

Compressed Institutional Time

There is a structural reason the region can close this chasm faster than the West. It is the same reason it could mistake adoption for arrival. The West built its institutions over a century, then spent thirty years digitising them, bolting technology onto plumbing laid in a different era. Every AI deployment in a legacy Western firm is a renovation: a core to migrate, a process to protect, an incumbent way of working defending its own obsolescence.

The Gulf often does not carry that legacy. It is building the institution and the technology at the same time. Call it Compressed Institutional Time: when you construct the institution and its enabling technology together, you do not run the same curve faster. You skip stages of it. The Hajj operation is Compressed Institutional Time made visible. The orchestration capability and the institution doing the orchestrating were built together, digital-native, with no analogue process to defend.

The same conditions that let you skip stages let you skip the unglamorous capability-building too, and report a high adoption number while the chasm stays wide open.

That is the regional edge, and it is also the trap. The advantage is a licence to build, not a guarantee of building.

Why Lag Is the Wrong Word

The strongest objection is simple. High adoption naturally precedes value, so the chasm is just lag. Buying comes before building in every technology cycle ever recorded. The 11% becomes the 40% becomes the majority. The gap closes itself. The sequencing is real, and I take it seriously. Adoption does come first. Nobody realises value from a tool they never bought. But first is not automatic.

The chasm closes only where someone does the deliberate work: the judgment hire, the process redesign, the operating-model change. Where that work is skipped, time does nothing. A firm can sit at 84% adoption and 0% value for years, because no mechanism rewires a business around a tool it never integrated. Idle software does not compound. Lag implies motion toward a destination. Much of the chasm is not motion. It is a parked tool and a slide that says AI-enabled.

The proof is in the numbers themselves. Watch how the three figures move together, or fail to.

If adoption alone pulled value behind it, those gaps would narrow in proportion. They do not. Adoption raced. Value did not follow. That is the signature of a gap that requires work to cross, not time to wait out.

The Strategic Point

Here is what the 84% headline hides. The race everyone thinks they are running is the adoption race: who deploys fastest, who touches the most functions, who claims the highest percentage on stage. That race is nearly over, and it never mattered much. The race that decides the decade is the chasm race: who converts adoption into capability, fastest and most deliberately.

The region that closes the chasm first wins. Not the one that adopts fastest. For the Gulf the implication is sharp. The Compressed Institutional Time advantage is real, and the Hajj proves the capability ceiling is high. But the 84% is the easy part, and the region has done it. The 11% is the hard part, and the region has the rare conditions to win it, if it treats the chasm as the work rather than a delay to be waited out.

Buy a tool and you have a line item. Build a capability and you have a moat. The chasm is not where you wait. It is where the decade is won.