Some would say that many Latin American public officials entertain a very seductive idea about digital sovereignty: that a law or decree is all it takes to impose it. Yet what sounds simple in words is anything but in the world of technology, especially when you examine the telecommunications network connected to the data center, the services it is expected to support, its customers, and its redundancy.
Every aspect becomes more complex when you consider that, in Latin America, this interwoven infrastructure and service ecosystem often crosses several national jurisdictions. Therefore, reducing digital sovereignty to requiring data to reside within national territory, reserving strategic infrastructure for the State, and distrusting every foreign provider as a way for a country to regain control over its technological destiny is an excellent recipe for failure.
Nevertheless, Latin American history is filled with leaders who have prioritized words over actions. Accordingly, promising what is politically appealing but impossible to implement has not stood in the way of introducing bills, drafting decrees, or making campaign promises. Add to this backdrop the weakness of many regional government institutions and the growing discretion exercised by those in power, and the result is not a promising outlook for infrastructure investment in certain jurisdictions. This comes at a time when the expanding use of artificial intelligence (AI) and the arrival of new technologies over the next ten years make it almost imperative to establish a formal dialogue among all stakeholders in order to reach an agreement that benefits everyone.
On the one hand, fears about the loss of sovereignty in an increasingly globalized world must be dispelled; on the other, investors must not perceive high levels of risk in their short-, medium-, and long-term investments.
Within this important debate, Argentina stands out as one of the regional markets with the longest track record on the issue. In 2000, for example, it enacted Law No. 25,326, one of Latin America’s first data-protection regimes. The European Union recognized the country as an adequate jurisdiction in 2003, a status reaffirmed in 2024.
The market also has digital signatures, access to public information, an enforcement authority, and State infrastructure centered on ARSAT. Argentina can therefore be said to have experience legislating on data. Some domestic sectors argue, however, that the country has failed to enforce those rules with a deterrent effect, capture the economic value of that information, or control data flows as they cross the border.
According to the Digital Sovereignty Index (DSI), which is linked to the BRICS organization, Argentina scores 2.0 out of 5.0. The document points to a relatively mature regulatory framework but limited operational control capacity. Nevertheless, one could infer that, in reflecting the interests of BRICS+, the DSI places particular value on autonomy from Western technology providers, domestic infrastructure, and State control of data, while giving less weight to international cooperation, open markets, the protection of digital rights, or dependence on technologies originating in other blocs.
The DSI is the only existing international comparison on digital sovereignty. Regardless of its bias, it therefore offers a comparison of the international landscape from an apparently statist perspective and illustrates one of the main elements of Argentina’s digital-sovereignty debate.
On one side are the laws’ intended goals; on the other is what they actually achieve under that legal framework. Rules are one thing, while effective constraints are quite another: the real capacity to enforce, audit, sanction, and respond to an extraterritorial request. For both an investor and the country seeking to attract that investor, the formal ambition of a rule matters less than its enforcement, predictability, and economic effect. A sophisticated framework that no one enforces does not produce sovereignty; it produces uncertainty. And uncertainty is precisely the raw material of discretionary centralism.
It is worth clarifying what digital sovereignty means, because confusion gives rise to the worst public-policy mistakes. It is not the same as security: data may be perfectly encrypted yet not sovereign if the provider’s jurisdiction allows extraterritorial demands, such as those authorized by the U.S. CLOUD Act. Nor is it the same as autarky: installing a server on Argentine soil does not automatically place encryption keys, contracts, updates, or access decisions under national control.
Properly understood, sovereignty is the effective capacity to decide: to select and replace providers, impose verifiable controls, audit access, and migrate a workload when regulatory, contractual, or geopolitical conditions change. It is a dimension of resilience, not an exercise in State ownership.
This is where statist sovereignty rhetoric reveals its flaw. Faced with Argentina’s gap, the temptation is to close it with more State intervention: mandatory localization, discriminatory preferences, forced technology transfers, and discretionary controls over who may operate and under what conditions. The problem is that these tools require exactly what the country currently appears to lack: the institutional capacity to design them well, apply them impartially, and sustain them over time.
In the hands of a strong and predictable State, a localization policy can reduce certain jurisdictional risks in exchange for higher costs. In the hands of a weak State, the same policy becomes a lever for arbitrariness: rules applied at discretion depending on the government in office, the favored sector, or the political circumstances. The result is the worst of both worlds: arbitrary control without the capacity to exercise it. There is no sovereignty, only discretionary power disguised as sovereignty.
The economic cost of that disguise is real. The most mobile capital, capital that operates on global platforms and can move to another jurisdiction, flees discretion before it flees taxes. And capital-intensive projects with long payback periods, such as data centers, networks, or AI computing infrastructure, simply cannot be financed without contractual stability and clarity about exit conditions. A regime that can change the rules by decree does not attract long-term investment: it drives it away or makes it more expensive through a political-risk premium that the country itself ultimately pays in the form of less installed capacity.
In light of that diagnosis, a pro-investment approach does not mean opening the doors without conditions or handing critical infrastructure to whoever arrives first. It means replacing discretion with rules. Properly understood, that is the merit of the Incentive Regime for Large Investments (RIGI): its strength lies less in its tax benefits than in how it ties the State’s hands by offering tax, customs, and foreign-exchange stability for up to 30 years. Predictability is the incentive. A government that commits not to exercise discretion over a project is doing something more valuable than granting a subsidy: it is building the trust that its weak institutions cannot provide on their own.
As one would expect, consistency requires applying the same rigorous standard to good news. The project announced as Stargate Argentina (up to USD 25 billion and 500 MW of computing capacity in Patagonia, linked to OpenAI and Sur Energy) illustrates both the opportunity and the danger of confusing an announcement with a fact. For now, the project remains a letter of intent, with no binding contracts, definitive financing, or published timetable.
The picture is completed by the ‘Super RIGI,’ which seeks to extend benefits to hyperscale data centers and is still awaiting legislative approval, while the intention to bring private capital into ARSAT does not mean that its privatization has been approved. Distinguishing among a rule currently in force, a legislative proposal, an international commitment, and a campaign announcement is not a technicality: it is the first discipline of any serious investment decision. USD 25 billion is an announced ceiling, not a guaranteed investment.
It would be dishonest to ignore the sharpest criticism of the open-market model, which in Argentina has even been described as ‘techno-Caesarism’: the concentration in the hands of global private actors of functions once reserved for the State, actors that would not replace it by force but would render it unnecessary through dependency. The concern has merit. Installing servers does not guarantee sovereignty over the data they process; a 500 MW data center creates real energy and water demands; and a project with little local content may face legitimate opposition even if it complies with the letter of the law.
The point is that the remedy proposed by that criticism, more discretionary State control, worsens the disease in a country with fragile institutions. The correct response to dependency is not autarky, which would be impossible and ruinous for a medium-sized economy, but the verifiable management of unavoidable dependencies. That is achieved through rules, not political will: local custody of cryptographic keys, multicloud architectures that prevent lock-in to a single provider, genuine workload portability, continuous access auditing, and, above all, clarity about which jurisdiction applies to information hosted in the country.
The commitments that matter (talent development, competitive local procurement, energy efficiency, and access to computing capacity) are better negotiated in transparent contracts than imposed through discretionary preferences that ultimately protect inefficient providers.
In this area, Latin America has a model close at hand that it has failed to recognize. The Caribbean, compelled by its vulnerability to hurricanes and floods, is advancing the concept of data embassies: a mechanism grounded in principles of the Vienna Convention that allows critical copies of government information to be hosted outside the territory while remaining under the jurisdiction of the State of origin.
In December 2025, the Economic Commission for Latin America and the Caribbean (ECLAC) published a study on the subject, and in June 2026 twelve delegations endorsed the Belize City Declaration. This is sovereignty by treaty, not by decree: an agreed, precise, and enforceable legal constraint that separates territory, jurisdiction, and control. It is the exact opposite of discretionary centralism. Argentina could explore both this outbound model, as a complement to ARSAT, and an inbound model that clarifies the law applicable to foreign data hosted in the country, provided it carefully assesses which jurisdiction it would be willing to relinquish.
The operational question of digital sovereignty is not whether a country owns every layer of the infrastructure or uses exclusively domestic providers. It is whether companies and public agencies can decide where data is processed, who accesses it, how it is protected, and under what conditions it can be moved.
Investment in data centers provides the computing, storage, and connectivity capacity needed to train and operate advanced AI models. It also reduces dependence on foreign technology, attracts talent and investment, and strengthens the country’s digital sovereignty and competitiveness.
This is extremely important in an environment where the outlook indicates that the growth in AI use will drive, according to McKinsey, USD 7 trillion in global data-center investment by 2030. Together with the greater use of new advanced technologies, this will lead to annual growth of between 16% and 25% in the power capacity allocated to information and communications technology through that same year, according to Bain & Company.
The discussion is not simply about sovereignty, but about creating the conditions needed to make Argentina a major beneficiary of this investment. The country does not need a heavier discretionary hand from the State; it needs stronger institutions and more predictable rules. A State that limits its own discretion attracts capital and, paradoxically, retains more control than one that reserves the right to change the rules whenever convenient. That is the real shortcut to sovereignty: not the decree, but the rule that binds even the person who issues it.
