From my view, Argen
tina is making a dangerous bet with artificial intelligence.
On June 4, 2026, Buenos Aires Times reported that President Javier Milei promised tech firms new laws and “unregulated” AI in Argentina.
The offer was direct: bring AI business to Argentina, and the government will not slow you down with heavy rules.
Milei made this argument in a Financial Times opinion piece written with Federico Sturzenegger, Argentina’s Deregulation & State Transformation Minister. They said AI needs a new legal structure. One idea is a “non-human corporation,” meaning a company run by AI agents or robots instead of human managers.
That sounds bold on paper. But one question stands in the way:
Who is responsible when AI causes harm?
That question must be answered first.
“Unregulated AI” is more than a business idea. It is a legal risk.
Milei compared today’s AI moment to 1602, when the Dutch East India Company helped spread the limited liability company model.
His point is that new technology needs new company law.
I agree that AI needs new law.
But without responsibility?
If an AI-run company acts on its own, the law must say who answers.
Is it the investor?
The programmer?
The company owner?
The AI provider?
The user?
The government that allowed it?
If the answer is unclear, the system is not ready.
Argentina can attract AI firms, support data centers, software companies, and new digital industries. But before any such law, Argentina needs five hard rules.
- Human responsibility. Every AI-run company must have a real human or legal owner responsible for its actions. AI cannot become a shield for avoiding blame.
- Data protection. Companies must clearly state what data they use, where it comes from, and whether people can refuse. Private data should not become free fuel for AI systems.
- People should know when AI is making or helping make an important decision about them.
- Risk control. AI used in health, finance, jobs, education, public services, courts, policing, or elections must face stronger rules than simple low-risk tools.
- If AI causes financial, legal, medical, or personal harm, someone must pay. A machine cannot be sued in any meaningful way unless real people or companies stand behind it.
This does not block AI. It is basic public protection.
The OECD AI Principles, adopted in 2019 and updated in 2024, call for trustworthy AI based on human rights, safety, transparency, and accountability.
The European Union AI Act, which entered into force on August 1, 2024, uses a risk-based model. It does not ban AI. It puts stronger rules on higher-risk systems.
That is a safer direction.
Argentina does not need to copy Europe. But it should not sell “no rules” as a business plan.
The danger is not only private business. In May 2026, Argentina also discussed a “social digital twin” program to help create public policy with AI.
When AI enters government, the danger grows.
A bad private AI system may hurt customers.
A bad public AI system may hurt citizens.
That is not innovation. That is legal escape.
From my view, unregulated AI in Argentina is not freedom.
It is a red flag.
Technology can move fast.
Law must not sleep.