McKinsey study: AI agents and robots could reshape half of Latin America's work hours

A new study from McKinsey Global Institute (MGI), the economic research arm of McKinsey & Company, asks a question that until recently seemed reserved for the world's wealthiest economies: how much of the daily work in Latin America could a machine already do? The study, titled "Agents, robots, and us: How AI reshapes work and skills in Latin America," gives a striking answer —more than half of today's work hours in the region, around 57%, could technically be automated using artificial intelligence and other existing technologies—, arriving at a moment when businesses from Houston to Bogotá, and from Monterrey to Panama City, are deciding, at different speeds, how to bring AI into their daily operations.

An automation potential on par with advanced economies

According to McKinsey, the share of technically automatable work hours in Latin America is comparable to what the firm has calculated for advanced economies in prior research on the United States and Europe. That finding cuts against the intuition that AI-driven automation is mostly a rich-country phenomenon tied to high wages and expensive labor. McKinsey's analysis covers 15 Latin American countries and concludes that, technically, the region has as much to gain —or reorganize— as any developed economy.

That technical potential, however, is not the same as actual adoption. McKinsey draws a clear line between what technology can do today and what companies are actually doing with it, a distinction that is central to understanding why a figure as large as 57% has not yet translated into an immediate wave of layoffs or mass restructuring across the region.

Why AI agents, not robots, concentrate the value

One of the study's more specific findings concerns which type of technology does the heavy lifting within that automation potential. According to McKinsey, AI agents —software capable of executing tasks, making decisions within defined rules, and operating with relative autonomy— account for most of the potential automation value in Latin America, similar to what happens in other regions of the world. In the analysis's midpoint scenario, agents account for about 80% of that value.

Physical robots, by contrast, play a comparatively larger role in Latin America than in advanced economies, precisely because the region concentrates a higher share of physical and manual labor within its economy. It's an important distinction: while a country with a more developed services economy tends to see automation concentrate in offices and digital workflows, in Latin America AI software and physical robotics are advancing in parallel, each addressing a different segment of the labor market.

What explains the region's likely slower pace of adoption

McKinsey is just as specific about why actual adoption of these technologies in Latin America is likely to move slower than in advanced economies, even if the technical potential is comparable. The firm points to three concrete factors: lower wages in the region, which reduce the immediate economic incentive to replace cheap labor with technology; the relatively higher cost of robotics compared to those wages; and differences in companies' organizational readiness to absorb change at this scale.

This combination of factors helps explain why such a large finding —more than half of work technically automatable— does not equal an immediate transformation. The gap between what technology allows and what organizations actually implement is, according to McKinsey, the real battleground of the coming years, more so than the availability of the technology itself.

At the same time, the study documents that demand for AI-related skills in the region's labor markets —where job-posting data is available— is growing at roughly twice the rate seen in advanced economies. That is a signal that even as technology adoption proceeds cautiously at the organizational level, companies in the region are already incorporating AI into their day-to-day operations enough to actively seek out that talent in the market.

A technical potential, not a layoff sentence

McKinsey is explicit on a point that often gets lost in coverage of studies like this one: the percentage of automatable hours should not be interpreted as a forecast of how many jobs will disappear. It is, according to the firm, a technical potential —what technology allows today—, not a prediction of what companies will actually do, nor of how many people will lose their jobs as a result.

The reading McKinsey proposes instead is different: this number is, above all, a signal of how deeply work processes could be reorganized in the coming years, as companies combine human workers, AI agents, and robots to expand coverage, speed up decisions, and take on tasks previously considered impractical to automate. It's a framework of "collaboration" rather than "replacement" —though McKinsey also acknowledges that how each company combines those three elements will determine how real that projected $450 billion in value actually turns out to be.

What this means for your business

While McKinsey quantifies this automation potential at a macroeconomic level, for an individual business in Houston, Cypress, Mexico City, or Bogotá, the question translates into something far more immediate: if your customers are already using AI agents to search, compare, and decide, does your business show up in that conversation? The study focuses on how AI transforms companies' internal operations, but the inverse question is just as urgent for any business that depends on being found: the same AI agents and answer engines McKinsey describes transforming work are also changing how consumers discover and choose businesses every day.

That's exactly where MerchandisePROS's AI Search Optimization (AEO) service comes in. If AI skills demand is growing at twice the regional rate, per McKinsey, it's reasonable to assume that consumer use of AI assistants to make purchasing decisions is growing at a comparable pace. Our audit evaluates exactly the signals that determine whether ChatGPT, Perplexity, or Google's AI Overviews recommend your business with confidence, and delivers a concrete plan to close that gap before your competitors do. You can review the full detail of this and other services on our services page.

"McKinsey's study confirms something we see every day working with businesses across Latin America and the United States: the technology to automate work is already here, but the real edge goes to whoever also makes sure AI recommends them, not their competitor."
- Diego Medina F, Founder of MerchandisePROS

Frequently Asked Questions