Milei vs. Harari: AI-Owned Companies? The Madness to Come

President Javier Milei has stirred up a hornet’s nest. In an article published in the Financial Times, he defended the creation of a new legal category: «automated societies,» companies that would operate through algorithms or artificial intelligence agents, without the need for human beings. And not only that: he also proposed granting legal personality to DAOs (decentralized autonomous organizations). The goal, he says, is to position Argentina as a hub for technological innovation and attract investments. To that end, the Executive Branch has already submitted a bill that completely reforms the General Companies Law.

But Israeli historian and writer Yuval Noah Harari responded, also in the Financial Times, with a bombshell: granting legal personality to AI would be like handing over a «master key» that would allow non-human agents to access financial, economic, and political systems. Harari warned that these companies could exploit legal loopholes and that, since sanctions like prison cannot be applied to an AI, a vacuum in accountability would open up.

This debate may seem distant for those juggling to pay their credit card or health insurance, but it has a direct impact on the life of any ordinary citizen. Let’s break it down.

Today, our Civil and Commercial Code distinguishes between natural persons (flesh and blood) and legal persons (any corporate organization). Both are «subjects of law,» meaning they can acquire rights and incur obligations. The law recognizes attributes such as name, domicile, and assets. Thus, you are a natural person; Arcor Sociedad Anónima, a legal person.

The Argentine legislator created different types of corporate entities as a «legal fiction» so that partners can carry out an activity using assets separate from their own. If the business fails, the partners’ liability is limited, in principle, to the capital they contributed. For example, if with two friends we open a bar in Palermo and create «El Bar de Fernando SA» with a capital of 50 million, if the bar is a success we share profits; if it fails, we lose the capital. But if the company causes harm to third parties, breaches contracts, evades taxes, or harms workers, the legal system allows sanctioning those actually responsible: administrators and even partners. The company does not think or decide; its partners, administrators, directors, or representatives do.

The Executive’s bill proposes creating, on one hand, companies capable of operating through autonomous algorithms or AI agents without workers or human resources, and on the other, blockchain-based DAOs. Here comes the million-dollar question: who is responsible when the algorithm makes a mistake and causes harm? AI can execute decisions, but it cannot assume legal responsibility. It cannot be sued, sanctioned, or appear before a court. It is clear that behind every automated decision, there must still be concrete individuals responsible, both civilly and criminally.

Furthermore, there is a risk of erosion of corporate governance principles: transparency, accountability, and the conduct of administrators. Replacing administrative bodies with algorithmic processes can create gray areas incompatible with the control requirements of a complex economy. International experience shows that DAOs face severe governance problems.

It is not about being anti-technology, but innovation should not be an excuse to weaken principles that have proven effective for decades. Responsibility cannot be delegated to an algorithm. And if the President insists, he should prepare for the debate: Harari has already put the cat among the pigeons.

WATCH THE VIDEO: