Something shifted in October 2025 that most people in the tech industry underestimated.
OpenAI and Argentine energy firm Sur Energy signed a letter of intent for "Stargate Argentina", a proposed AI data center in Patagonia with up to 500 megawatts of capacity, powered by renewables, representing a potential investment of up to $25 billion. The project is being led by Sur Energy, with OpenAI as a potential anchor tenant. OpenAI CEO Sam Altman called it "more than just infrastructure," describing it as "putting artificial intelligence in the hands of people across Argentina." It would be the first Stargate-linked project in Latin America. The region chosen wasn't Singapore, wasn't Dubai, wasn't Warsaw. It was Patagonia.
That announcement didn't happen in a vacuum.
A continent-wide infrastructure bet
In the months before and after, the signals kept coming. AWS committed $5 billion over 15 years to Mexican cloud infrastructure, launching its first cloud region in the country in January 2025. Microsoft announced $1.3 billion over three years to expand AI infrastructure and skills programs in Mexico, announced personally by CEO Satya Nadella at the company's AI Tour in Mexico City. Brazil's government launched a dedicated national AI plan allocating approximately $4 billion over four years. Actis launched TERRANOVA, a new hyperscale data center platform, with plans to invest $1.5 billion across Brazil, Mexico, and Chile over three years.
These are not soft commitments or brand signals. These are infrastructure decisions, the kind where companies stake real capital on where they believe AI will be built, trained and consumed in the next decade. And right now, those bets are pointing squarely at Latin America.
The reasons aren't hard to find. According to Brazil's own National Energy Balance, 88% of the country's electricity came from renewable sources in 2024, one of the cleanest grids of any major economy in the world and an increasingly decisive advantage as AI data centers become some of the largest consumers of power on the planet. Chile's cool climate and renewable energy access have made Santiago one of the most ESG-friendly data center hubs in the region. Mexico's Querétaro cluster has grown to over 160 megawatts of capacity, becoming a dominant digital infrastructure zone driven in large part by nearshoring demand.
The talent gap no one is talking about
Here's the paradox at the heart of this moment. Latin America accounts for a meaningful share of global GDP but receives a disproportionately small share of global AI investment. The infrastructure is arriving. The adoption is happening, some analyses suggest Latin America may be overrepresented in global AI tool usage relative to its share of internet users. But the local capital and institutional support to build AI from scratch still lags far behind.
Some analyses suggest Latin America may be overrepresented in global AI tool usage relative to its share of internet users.
What that creates, for US companies paying attention, is an asymmetric opportunity. The engineers are there, trained on modern stacks, working in overlapping time zones, embedded in an ecosystem that is being wired, right now, with the same AI infrastructure that powers Silicon Valley. But the competition to hire them is still a fraction of what it is domestically.
In one regional hiring report they state that software engineering placements in the region grew 250% year over year, with 84% of those hires being mid-level and senior professionals. This isn't companies offshoring support tasks. This is engineering leadership being built in Latin America.
What the ILIA 2025 index actually tells us
Every year, ECLAC and Chile's National Center for Artificial Intelligence publish the Latin American AI Index, the most rigorous country-by-country measurement of AI readiness in the region. The 2025 edition draws a clear map.
Chile, Brazil and Uruguay score above 60 and qualify as "pioneers", countries with consolidated infrastructure, specialized talent pipelines and functional AI governance frameworks. Colombia, Costa Rica, Ecuador and others sit in the "adopter" tier, having closed the gap significantly through improvements in connectivity and national strategy.
The report also notes a strong resurgence of data center investment across the region, with multi-billion-dollar private sector announcements over the next decade.
The geography of AI readiness in the region is clearer than it's ever been and it maps directly onto where the world's best nearshore engineering talent sits.
What this means for US companies building AI products
The hyperscalers are betting on a region in the middle of a rapid upgrade, in compute, in connectivity, in engineering sophistication and in government support for the technology sector.
For US companies building AI products, the question is no longer whether Latin American engineers can contribute meaningfully to frontier work. The question is whether you want to access that talent before or after your competitors do and whether you'd rather build those relationships when the market is still relatively uncrowded, or in three years when projects like Stargate Argentina are operational and every major tech firm has realized what OpenAI already has.
At South Geeks, we've been building nearshore engineering teams from Latin America for years. What we're seeing in 2025 and 2026 isn't an acceleration of an old trend. It's the beginning of a different era: one where the region stops being a cost play and becomes a strategic one.
South Geeks connects US companies with senior engineering talent from Latin America. If you're building an AI product team and want to understand what's actually available in the region right now, let's have a conversation