Milei Wants Argentina to Be the First Home for AI-Run Firms

Politics

Argentina is trying to become the first country in the world to let a non-human corporation exist in law, a company that an artificial-intelligence system rather than a person would run. Read the draft, though, and the machine still needs a human standing behind it.

President Javier Milei sent the bill to Congress on May 29, proposing to rewrite the General Companies Act that has governed Argentine business since 1972. The headline idea is a brand-new legal category, the “automated company,” that could operate through algorithms or AI agents, with human shareholders allowed but no longer required.

Milei set out the logic in a Financial Times opinion piece written with his deregulation minister, Federico Sturzenegger. He reached back to the 1602 founding of the Dutch East India Company to argue that limited liability was the legal breakthrough of the age of sail, and that the AI era now needs its own version.

The rest of his message was blunter. “We are open for business,” he declared, pitching Argentina as a place where AI firms would face low corporate tax and a promise to leave the technology largely unregulated.

Why the non-human corporation still needs a human

The branding runs ahead of the text. According to a July analysis by Reuters, the automated company would be required to keep a human administrator to oversee operations, and the firm would be liable for damage caused by its AI or algorithmic systems.

The most autonomous format in the draft, a blockchain-based structure modelled on a decentralised autonomous organisation, leans on people too. Legal analyses of the bill note it still needs a human legal representative to sign binding acts, plus a human promoter who answers without limit for the company’s obligations at formation.

Where anti-money-laundering rules apply, a human compliance officer is required as well. In short, the accountable person never fully disappears, which matters for any foreign investor weighing how much real risk shielding the structure would offer.

Diego Duprat, a law professor who co-wrote the text, argued that automated businesses already exist in limited forms, pointing to cashier-less supermarkets. Lawrence Cunningham of the University of Delaware called the proposal bold but modest in practice, saying it recognises you might run a business without any human resources department rather than reinventing the company itself.

A bid to become the Delaware of AI

The commercial logic is a package deal aimed at one customer: the founder building autonomous, agent-based businesses. Argentina is offering a bespoke legal container, limited liability, low taxes and a hands-off regulator, hoping to attract company formations the way American states like Delaware attract incorporations.

Foreign minister Pablo Quirno and the wider Milei team have tied the idea to a broader campaign to sell Argentina as an AI base, citing cheap Patagonian power and cold weather suited to data centres. It sits alongside the “Super RIGI” incentive bill, which offers thirty years of tax and currency stability to large new projects.

The timing is deliberately contrarian. The United States, the European Union and China are all tightening AI oversight, so Argentina is betting that a lighter regime can pull in capital that stricter jurisdictions may push away.

Investors reading from London or Munich should treat the signal carefully. One Silicon Valley venture partner told Reuters that regulation alone will not turn the country into a hub, and that the deciding factor is whether AI talent actually relocates there.

The accountability fight the non-human corporation has started

The proposal drew a sharp international response. In a counter op-ed, historian Yuval Noah Harari warned that AI-run firms could become expert in regulatory arbitrage, and that the ultimate deterrent for human executives, prison, means nothing to software.

Harari turned Milei’s own comparison against him, noting that the Dutch East India Company burned the port of Jayakarta in 1619 and ruled as a private empire. Buenos Aires, he cautioned, risked becoming a new Batavia rather than a new Amsterdam.

Microsoft AI chief Mustafa Suleyman agreed, writing that AI agents deserve no more legal standing than a laptop. Milei replied on social media that giving these entities a defined legal category would make them easier to regulate, not harder, since authorities could point at a named structure instead of software operating in the shadows.

What is a non-human corporation under Milei’s bill?

It is a proposed new legal category, called an automated company, that could operate through artificial intelligence or algorithms rather than requiring human owners or staff. The draft still requires a human administrator to oversee the firm, and the company would be legally liable for damage its AI systems cause.

Has Argentina passed the law yet?

Not yet, because the measure was submitted to Congress on May 29, 2026 and remains a draft that has to be debated and voted on. It would replace the General Companies Act that has been in force since 1972, so it is a full rewrite rather than a minor tweak.

Why does the plan matter to foreign investors?

Argentina is trying to position itself as a low-tax, light-regulation base for AI companies at a moment when larger economies are tightening their rules. For investors, the open questions are how much liability protection the structure truly offers and whether AI talent and capital actually move to the country.

Frequently Asked Questions

What exactly is the 'automated company' proposed in Argentina's draft bill?

An 'automated company' is a new legal category proposed by President Javier Milei that would allow a firm to be run by algorithms or AI agents, with human shareholders permitted but no longer required. The bill was sent to Congress on May 29 and would replace Argentina's General Companies Act, which has been in force since 1972.

Does the proposed law truly remove all human involvement from these AI-run companies?

No, despite the 'non-human' branding, the draft still requires a human administrator to oversee the firm. The company itself would also be held liable for any damage its AI causes.

Who has criticized the proposal and what are their concerns?

Historian Yuval Noah Harari and Microsoft AI chief Mustafa Suleyman have warned that the idea removes human accountability and invites regulatory arbitrage. The measure remains a draft in Congress and has not yet become law, with officials confirming no firms or investment commitments are currently tied to it.