Argentina's Congress weighs bill allowing companies to operate without employees

Critics warn bill would blur accountability for AI-made decisions

The proposal would create two new corporate forms: automated companies, which could conduct activities through “autonomous algorithmic systems or artificial intelligence agents” and operate “without requiring employees or human resources for its ordinary operations,” according to the bill’s text, and DAOs, which make collective decisions through rules programmed into computer systems. The government said the changes are intended to adapt the law to technologies that can perform tasks and make certain decisions that have traditionally depended on people.

In supporting arguments reported by Argentine financial newspaper Ámbito Financiero, the government said technological changes are creating “new activities, new markets, new sources of financing and new forms of governance.” It maintained that these forms of organization require specific recognition under Argentine law.

The proposal arrives after Milei used his address to the U.N. General Assembly to oppose preventive regulation of artificial intelligence, arguing that government control over innovation gives bureaucrats the power to restrict human creativity. Milei advocated a less restrictive approach to AI and aligned his position with the United States’ market-oriented approach. He pledged a framework emphasizing limited government intervention, tax incentives and limited liability to attract global investment in the sector.

Maximiliano Firtman, an artificial intelligence expert, challenged the central premise of the bill. “To me, Milei’s idea is delusional. It is not based on the reality of how technology works,” Firtman said. He argued that artificial intelligence executes instructions established by people and has no will of its own, and that for that reason an AI system cannot independently assume the role of a manager or the legal obligations assigned to a person.

“The fact that companies increasingly automate tasks with artificial intelligence agents does not mean they can do without people,” Firtman said. “There will continue to be humans and AI agents, each with different functions.”

Firtman identified one of the proposal’s main risks as determining responsibility when a decision made by an artificial intelligence system causes harm. He said that if a company does not have a clearly identified person responsible for the decision, holding someone accountable could become more difficult. He also warned that a lack of clarity could make it easier to use such companies to conceal the people behind illegal activities, citing fraud and money laundering as potential risks.

Pablo Serdán, an attorney specializing in artificial intelligence and corporate power, raised similar concerns about the structures’ potential to limit legal liability. “What they want to implement in Argentina is an experimental field where companies can come in, have algorithms make their business decisions and reduce their liability if they cause harm,” Serdán said.

Serdán said the proposal does not establish a minimum capital requirement or necessarily require a physical address in Argentina, provisions he said could make it more difficult to identify who is responsible when a company causes harm, particularly when decisions rely on automated systems.

Serdán raised broader concerns about DAOs. He said such organizations can be used to manage resources and make collective decisions through rules programmed into computer systems, and warned that granting them legal recognition could expand their use into areas that do not necessarily involve commercial activity. “It introduces new risks to our democracy,” Serdán said.

According to Argentine business publication iProfesional, the proposal is based on academic work by Emiliano Kargieman, CEO of Satellogic, a company that provides satellite imagery services. The bill remains under consideration in Congress and could be amended during the legislative process. If approved, Argentina would add the two new corporate structures to its company law and establish a legal framework for businesses that use autonomous systems and artificial intelligence agents as a central part of their operations.