アルゼンチン、AI所有の企業設立を提案

BUENOS AIRES, ARGENTINA - MAY 22: President of Argentina Javier Milei interacts with supporters after the presentation of his new book 'Capitalismo, Socialismo y La Trampa Neoclasica' (in English: Capitalism, Socialism, and the Neoclassical Trap) at Luna Park on May 22, 2024 in Buenos Aires, Argentina. The presidential Spokesperson Manuel Adorni announced that Milei will perform in a musical show. (Photo by Marcelo Endelli/Getty Images)Getty ImagesImagine a company with no CEO, board or human employees — just artificial intelligence making every decision, signing contracts and owning assets in its own name. That scenario once felt safely hypothetical — ripe subject matter for SciFi movies and novels for years.Argentina’s president Javier Milei announced last week that his government submitted legislation to “non-human corporations” — a legal entity owned and operated completely by AI agents or robots. “Human shareholders may participate, but are not required,” he wrote in a guest column in the Financial Times, co-written with Deregulation Minister Federico Sturzenegger. It set off a firestorm discourse.What exactly is President Milei proposing?The legislation rests on three pillars: no regulation of AI (“free to be developed without the deadly hand of premature and poorly understood regulation,” he stated in his column), a new corporate category for AI-run companies and a low corporate tax rate to lure tech investment to Buenos Aires. Shareholders would be able to choose their form of corporate governance law, although final beneficiaries would have to be disclosed.This “non-human corporation” bit is one of the most interesting parts. Under existing laws everywhere, a company needs human beings — directors, shareholders, someone legally accountable. In his column, Milei is essentially stating he wants to strip that requirement away, and in theory, that means an AI agent could incorporate a company, execute contracts, hire workers and sue people in court — all without a single human being making the call.Milei posited this as the natural next step after the limited liability company, which the Dutch East India Company pioneered in 1602.“The logic of 1602 still applies today,” he wrote in the FT.MORE FOR YOU “As much as the industrial revolution freed us from the constraints of the human muscle, AI will free us from the constraints of the human brain, pushing productivity beyond our wildest dreams.”One important caveat is that the actual bill before Argentina’s Congress — a larger investment incentive package known as “Super RIGI” targeting $1 billion-plus projects in sectors including AI data centres — does not itself mention non-human corporations. Hacking the game environmentFour days after Milei’s column ran, historian and Sapiens author Yuval Noah Harari published a direct rebuttal in the same newspaper, bringing up important questions: who do you punish when an AI-run company does something wrong? Who’s accountable?With human executives, there’s an easy answer, because they fear prison, and that fear — the self-interest of a person who doesn’t want to spend decades behind bars — is one of the things that keep corporate behavior within bounds. On the other hand, “it is unclear what kind of sanctions could keep it in check,” Harari wrote of an AI CEO. “If it faces bankruptcy — which is equivalent to its death — it would presumably be willing to do anything to avoid that fate.”Harari also cited a 2025 study by Berkeley non-profit Palisade Research in which advanced AI models from both OpenAI and DeepSeek, when playing chess against a powerful engine they were likely to lose to, frequently chose to cheat — hacking the game environment rather than accepting defeat. “By hacking the game environment, they could alter the result in their favour. Now imagine that the “game” is corporate competition, and the “game environment” is your country.” To elaborate on his point further, Harari drew on Milei’s own Dutch East India Company analogy: that company did pioneer limited liability and help make Amsterdam the financial capital of the 17th century, but its most consequential actions happened elsewhere — when it captured the port of Jayakarta in 1619, it burned the city down and built a colonial headquarters called Batavia, ruling the surrounding region as a private empire for its shareholders’ benefit. “Milei hopes to turn Buenos Aires into a new Amsterdam,” Harari concluded. “He risks turning it into a new Batavia instead.”The big picture playThe Milei government’s AI ambitions are part of a wider pivot to position Argentina as a destination for technology investment following a severe economic crisis. There has been growing interest from tech investors around the world, including billionaire Peter Thiel, who, according to the New York Times, recently spent extended periods in Buenos Aires, purchased property in the city, temporarily relocated his family there and met with Milei and senior government officials.Countries that grant AI legal personhood risk becoming something for which the historical record offers no analogy, argues Harari. “Not a company state, but an AI state — a country whose people could in effect be ruled by non-human corporations, against which it might be even more difficult to rebel.”Milei responded to Harari on social media, thanking him for joining this “fascinating and transcendental debate” and promising a formal reply. “Already preparing my reply to see if I we can appease your fears about the path I proposed last week.”

※ 著作権に配慮し、引用は冒頭3段落までです。続きは元記事をご覧ください。