Chile Won the AI Race. It's Still Falling Short on Impact.

There’s a number that should bother us more than it does. Chile has led artificial intelligence adoption in Latin America for three years running. We top the regional index, more than 80% of large companies already use it in some process, and the boardroom conversation has moved from whether it’s worth it to when we start. Any way you look at it, we got there first.

And yet only 3.6% of Chilean companies have managed to scale their AI projects with measurable returns. The rest, almost 70%, are still exploring or stuck in pilots that never quite take off. The figure comes from a study by MAS Analytics of 166 executives, and it’s worth pausing on, because it’s easy to misread.

Having AI isn’t capturing value

Having artificial intelligence and having artificial intelligence that’s useful are two different things. Buying licenses, opening accounts, asking a chatbot to draft emails: that’s having the technology, and Chile is ahead on that. Turning it into fewer wasted hours, processes that stop getting stuck, decisions that don’t wait until Monday: that’s capturing value, and that’s where we’re standing at the door.

What’s holding Chilean companies back

The interesting part is why. When those companies are asked what holds them back, almost nobody mentions technology or budget. Budget shows up as a barrier in just 1.2% of cases. What comes out on top is something else: data governance, organizational culture, resistance to change, the difficulty of connecting areas that never talked to each other. The study’s own director says it plainly: scaling doesn’t mean buying one more tool, it means redesigning processes.

And to be fair, this isn’t only a Chilean problem. Worldwide, the share of companies that have managed to scale artificial intelligence is barely a third. Nobody has solved this yet. The difference is that we got to the front of the line first and are now waiting at the same door as everyone else.

The leap happens at the bottom, not the top

Here’s a conviction, the kind you form by getting into companies’ real processes rather than reading about them in a report. The leap almost never happens from the top. Most AI plans start with a grand strategy, a corporate tool, a committee that meets every two weeks. And they stay there. The leap happens when the opposite occurs: when the tool reaches the hands of the person who does the work every day, and that person is given permission to redesign how they do it.

Something happens when that occurs. The person stops wondering what big AI project the company should take on, and starts seeing, in their own week, the specific places where time goes into something a machine does better. The report they put together by hand every Friday. The hour-long meeting that exists only to bring everyone up to date. The email they answer the same way twenty times. These aren’t revolutions. They’re small, specific frictions that this person knows better than any outside consultant.

And when those frictions are solved one at a time, with people inside and not from outside, they stop being an innovation expense line and start showing up where it matters: in results. Hours recovered. Processes that no longer need three people. That 3.6% that does capture value doesn’t do it because it has better technology than the rest. It does it because it stopped treating artificial intelligence as a technology project and started treating it as a redesign of how its people work.

The scarce resource isn’t the model

The technology, in all this, is almost beside the point. It’s available, it’s getting cheaper, and tomorrow it will be better than today. What’s scarce isn’t the model. It’s the decision to get into the slow, unglamorous work of looking at real workflows, one by one, and rebuilding them together with the people who live them.

We won the adoption race. It was a good first half. But the scoreboard that will matter in the coming years isn’t how many companies have artificial intelligence, but how many turned it into something you can notice. Nobody has won that race yet. And for the first time in a long time, Chile is in a position to get there first. The question is whether we’re going to run it seriously, or stay celebrating the first half.