To regulate or not to regulate AI? The debate is set to begin in Argentina's congress

Argentina’s congress will begin debating the regulation of artificial intelligence in the coming week.

A first hearing will be held on Tuesday to discuss a series of bills that aim to provide a regulatory framework for AI, partly inspired by European regulation.

Alarm bells can be heard among the business sector over bills that could reduce the country's competitiveness at a time when Argentina seeks to position itself as a center for innovation in AI, promoted by President Javier Milei.

“There is a contradiction between a rigid regulatory system and the dynamics of companies,” said Luis Galeazzi, executive director of the Argencon business chamber during a press conference.

“It is practically impossible to establish general rules [for AI],” he added.

According to Argencon, if AI regulation advances, rigid regulatory formats with an excessive bureaucratic burden, such as the European one, should be avoided.

The European law “is extremely extensive, complex to apply in practice, and very onerous for innovators and entrepreneurs,” the director of the technology and policy legal area at the Ecija law firm, Daniel Rodríguez Maffioli, previously told BNamericas.

The national government agrees that European “overregulation” is killing innovation.

However, Andrés Piazza, director of IDD-Lac, says there is no such dichotomy between innovation and investment. It is still early to understand the market implications of the regulation that Europe approved this year.

“Europe symbolically constitutes itself as a regulator. And, in addition, it has lower layers such as data governance regulations, digital services, digital markets, privacy and cybersecurity. It is a pyramid on which they build AI regulation,” Piazza said in statements sent to BNamericas.

Argentina does not have these types of regulations but it has laws on personal data, intellectual property, fraud, consumer protection and human rights, among others, that could apply to AI. Argencon says it is only necessary to adopt these regulations – or maybe adapt them – to the problems that may arise in relation to the use of AI.

“AI requires a regulatory system that understands the specificity of particular cases,” according to the business chamber. Argencon believes it is better to regulate by use case instead of having a general standard, as some bills due to be debated propose.

Galeazzi also valued the creation of regulatory sandboxes “in complex situations in which the implications of the use of AI in a particular environment must be resolved.”

Piazza said the most important thing about the discussion that is opening up in the country is to cover all sectors in order to have a state policy. “If not, what will surely happen is that we’ll have a law that does not understand the problem,” he said.

SECURITY

In recent days, the government advanced the use of technology within the State for security applications, one of the AI aspects that poses a high risk for the European Union.

An AI unit devoted to security (UIAAS) was created under the direction of the cybercrime and cyber affairs area of the security ministry.

The UIAAS will be in charge of patrolling social networks, applications and internet sites, as well as the dark web for the investigation of crimes, identification of their perpetrators and detection of situations of serious risk to security.

In turn, it will be able to analyze security camera images in real time to detect suspicious activity, and use machine learning algorithms to analyze historical crime data and thus predict future crimes and help prevent them, among others.