AI Infrastructure · Argentina · Paraguay · Energy · Data Centers

Argentina and Paraguay’s AI Infrastructure Bet: Can Energy Become Compute?

Argentina and Paraguay are trying to turn energy into AI and data-center relevance. The real question is not who announces the larger project, but who can convert power into operational infrastructure.

By Marcus A. Volz  ·  May 2026  ·  Econosur

The market signal

Argentina and Paraguay are no longer talking about artificial intelligence only as software, talent or public-sector modernization. Both countries are now trying to attach AI to physical infrastructure: energy supply, data centers, compute capacity and international technology partnerships.

That is the market signal. AI is making the physical layer of the digital economy visible again. The cloud is not abstract. It needs power, land, cooling, water, fiber, hardware, operators and long-term contracts.

For the Southern Cone, this creates a new infrastructure question: can countries with energy potential become relevant in the regional compute economy, or will most AI infrastructure continue to concentrate in established digital markets such as Brazil and Chile?

Paraguay: the clearer energy-to-compute story

Paraguay’s AI infrastructure argument begins with electricity. The country already has a strong renewable-energy identity built around hydroelectric power, especially Itaipu, one of the most important hydropower plants in the world.

In May 2026, Paraguay and Taiwan announced cooperation around artificial intelligence infrastructure. The Paraguayan presidency presented the agreement as a way to combine Taiwan’s semiconductor and technology capacity with Paraguay’s renewable and abundant energy potential.

This matters because Paraguay is not only announcing an abstract AI ambition. It already has a visible data-center layer. HIVE Digital has been expanding hydro-powered data-center capacity in Paraguay, including plans to increase its Yguazú site to 400 MW after an additional 100 MW expansion.

The country’s pitch is therefore relatively coherent: abundant renewable energy, lower operating-cost logic, existing data-center activity and a geopolitical technology link with Taiwan.

Argentina: the bigger but less proven ambition

Argentina’s AI infrastructure story is larger, more visible and more politically charged. OpenAI and Sur Energy signed a letter of intent to explore a large-scale data-center project in Argentina, with OpenAI describing Stargate Argentina as a potential first Stargate project in Latin America.

Reuters reported that the project could require investment of up to US$25 billion and involve a facility with capacity of up to 500 MW. That would place the project in a completely different scale category from Argentina’s current data-center base.

That contrast is important. CABASE, Argentina’s internet chamber, presented a 2026 infrastructure map showing only 13 operational data centers above 1 MW in Argentina, with approximately 32 MW of installed capacity, concentrated in the Buenos Aires metropolitan area.

Argentina therefore has two realities at the same time: a major AI infrastructure narrative and a comparatively small existing data-center base. The gap between those two realities is the core of the Argentine case.

Why Palantir is a signal, but not the proof

Peter Thiel’s visit to Argentina and his meeting with President Javier Milei add another layer to the story, because Thiel is associated with Palantir, data analytics, defense technology and state-scale information systems.

But this should be handled carefully. The meeting is a political and technological signal, not proof of data-center execution. Public reporting indicates that the meeting took place, while also noting that official details were limited and no confirmed data or intelligence services agreement had been announced.

For an Econosur reading, Palantir is useful as context: Argentina is trying to enter the global conversation around AI, data systems and strategic technology. But the hard infrastructure test remains elsewhere: megawatts, cooling, fiber, operators, regulation and contracts.

The real bottleneck is not the announcement

AI data-center announcements can look impressive because they use the language of national transformation: investment, technology, sovereignty, talent and regional leadership.

But data centers are execution-heavy infrastructure. Cheap or abundant power is necessary, but it is not enough. AI compute also requires reliable grid access, high-capacity transmission, cooling systems, water or alternative thermal design, redundant fiber routes, cloud partners, hardware supply chains and highly specialized operational teams.

That is where the comparison between Paraguay and Argentina becomes interesting. Paraguay may have a simpler value proposition, but it must prove technical depth and regional demand. Argentina may have a larger strategic narrative, but it must prove that a mega-project can move from announcement to physical capacity.

Energy is becoming a location factor

AI changes the value of energy. In older industrial strategies, cheap electricity was mainly connected to factories, mining, aluminium, chemicals or heavy manufacturing. In the AI economy, energy can also become a location factor for compute.

This does not mean that every energy-rich country becomes a data-center market. Compute infrastructure needs buyers. It needs latency logic. It needs regulatory stability. It needs talent and operations. It needs integration with regional cloud and enterprise demand.

But the strategic direction is clear: countries that can combine energy, connectivity and execution may gain a new role in the digital infrastructure map.

Paraguay vs Argentina: two different bets

Paraguay is betting on a compact infrastructure logic: renewable energy, lower operating cost, data-center operators and technology partnerships. Its advantage is clarity. Its challenge is to scale beyond energy into fiber, cooling, technical operations and regional cloud relevance.

Argentina is betting on scale and political visibility. The OpenAI-Sur Energy letter of intent, the RIGI investment framework and the broader technology narrative around the Milei government make Argentina visible to global investors. Its challenge is the large gap between existing capacity and announced ambition.

These are not the same strategy. Paraguay is trying to turn a narrow but strong advantage into a digital infrastructure position. Argentina is trying to use a larger national repositioning story to attract a step-change investment.

Who is this relevant for?

This trend is relevant for energy companies, fiber operators, cloud providers, industrial land developers, cooling specialists, data-center operators, equipment suppliers, construction firms and investors looking at South America’s next digital infrastructure layer.

It is also relevant for governments and regional development agencies. AI infrastructure is not only a technology topic. It touches energy policy, water management, tax incentives, foreign investment, telecom regulation and geopolitical alignment.

For companies evaluating Argentina or Paraguay, the practical question is not whether the countries have AI ambitions. The question is whether those ambitions create investable infrastructure ecosystems.

The execution checklist

For Paraguay, the decisive variables are whether renewable power can be paired with cooling, water management, fiber redundancy, operator competence, financing and international demand.

For Argentina, the decisive variables are whether large-scale announcements can pass through permitting, power contracting, grid integration, financing, cloud partnerships and long-term operational governance.

In both markets, the same principle applies: AI infrastructure only becomes real when the energy story meets the operational layer.