AI Regulation: Is Brazil Ready?

In an article published on the Estadão website, Dr. Matheus Puppe, partner in the TMT, Privacy and Data Protection area at Maneira Advogados, analyzes the debate surrounding the regulation of the use of artificial intelligence in Brazil.

AI Regulation: Is Brazil Ready?

07/10/2023

Matheus Puppe, Partner in the TMT, Privacy & Data Protection area at Maneira Advogados. Member of the CNJ Compliance GT and the Judiciary Integrity Committee (CINT). Guest professor at the postgraduate course in New Technologies, Compliance, ESG and Contracts at USP.

Disruptive technologies are expanding and the flow of global information is increasing at exorbitant speeds. Due to the constant evolution of technologies, we face new, inevitable daily challenges and constant changes. However, law as a crucial meta-technology for the fourth industrial revolution still thrives in a world increasingly based on algorithms, data and social accountability.

Law and, consequently, laws, adjust to new realities, framing the rights and duties of society, which is remodeled by new and progressive developments. As a mirror to life, law now reflects an interconnected, cosmopolitan and globalized world.

As a result, following the desires and needs of markets and societies per se, corporations, private entities, universities and even governments and individuals are in a constant process of creating global rules and laws. A process similar to the lex mercatoria, capable of effectively framing accelerated advances. An autopoietic law-making mechanism – now focusing on previously non-traditional parts.

Binding and self-regulating, the rule of law is reformulated to follow the new expectations of society, as well as protecting it by legislating on artificial intelligence, while observing new concerns and using technologies themselves to their advantage, in order to guarantee our rights globally, in an effective and innovative way.

Thinking about artificial intelligence certainly takes us to the future, but the problem is clear: the future has arrived. Science fiction is left aside when we observe that the process of canceling a telephone subscription is carried out primarily by robots, following the steps foreseen (or not) in their algorithms, dialoguing and, often, convincing the human user behind the phone call about the next steps and options.

Besides, who has never had to check “I’m not a robot” to access a specific website? Well, the fear of robots acting in a connected environment and increasingly interacting with human beings is now part of our daily lives.

But would the State not only have the power, but the capacity – technical and jurisdictional – to regulate robots? I believe so, and we are already moving in that direction.

Brazilian bills

In the last project presented, the law proposes protection measures for vulnerable groups, such as children, adolescents and people with disabilities, preventing companies from taking advantage of these vulnerable conditions, and establishes the creation of a body linked to the Presidency of the Republic to monitor the activities of AI companies in Brazil and monitor compliance.

One of the most important aspects of the bill are punitive measures for companies that violate the rules. The sanctions include fines of up to R$50 million for each infraction, prohibition of participation in experimental innovation projects (regulatory sandbox) for five years, suspension of the company's activities and prohibition of processing certain databases.

The project also guarantees rights to consumers, such as the right to be informed in advance of interactions with AI systems, to have explanations about decisions made by AI and the right to non-discrimination and correction of discriminatory biases.

A project very similar to the European initiative, but we need something legitimately Brazilian and adapted to our reality.

European Proposal

To keep Europe fit for the Digital Decade, the European Union has been accepting the new reality of artificial intelligence. In this way, they proposed a new regulatory proposal that aligns with the principles of data protection (GDPR), even giving the European Data Protection Authority the power to impose fines.

To this end, the European Union has been discussing the rules for the operation of Artificial Intelligence in both the public and private sectors, seeking to ensure the safety of people and fundamental rights (certainly a principled position). The EU's vision is for security, but also for international competitiveness, in order to facilitate innovation across the EU and to become a strong new player in the AI sector.

Some aspects stand out within this new type of regulation. For example, the proposal mentions that ""the same elements and techniques that enhance the socio-economic benefits of AI may also bring new risks or negative consequences for individuals or society. In light of the speed of technological change and potential challenges, the EU is committed to seeking a balanced approach […] Rules for AI available on the Union market or affecting people in the Union must therefore be human-centred so that people can trust that the technology is used safely and in accordance with the law, including respect for fundamental rights.”, clearly demonstrating its principled and flexible bias.

In this way, whenever AI systems are placed on the market or use initiated in non-compliance with the requirements of the regulation, Member States will have to establish effective, proportionate and dissuasive sanctions, including large administrative fines, in relation to infringements and immediate communication to the Commission.

The regulation provides some parameters for fines that must be taken into consideration.

Therefore, in order to harmonize national rules and practices in setting administrative fines, the Commission, with the advice of the Board, will develop specific guidelines, as carried out for the GDPR.

The Commission also intends to address the risks generated by specific uses of AI through a set of complementary and flexible rules (flexible even to align with the proposal for competitiveness in the global market). A structure that provides AI developers, implementers and users with the necessary clarity, having as its legal structure a clear and easy-to-understand approach, based on risk analysis, with four distinct levels of risk.

In addition, compliance is incorporated into the proposal by granting application providers that are not high risk the possibility of ensuring that their AI system is trustworthy, developing their own voluntary codes of conduct or adhering to the codes of conduct adopted by other representative associations. This will apply simultaneously with transparency obligations for certain AI systems. The Commission will encourage industry associations and other representative organizations to adopt voluntary codes of conduct.

The advancement of artificial intelligence (AI) in contemporary society is undeniable, and its influence extends to almost every aspect of our lives. Therefore, the need for robust regulation is urgent.

The recently proposed bill in Brazil, as well as the guidelines established by the European Union, are significant steps towards ensuring the ethical and responsible use of AI, and such initiatives highlight the importance of compliance with ethical AI, considering aspects such as transparency, privacy, non-discrimination and the responsibility of those who develop and implement these systems.

It is essential that such regulations establish robust protection measures for individuals and society as a whole, as well as promote the responsibility of companies that develop and use AI.

Despite this, regulating AI is not just a challenge for Brazil and the European Union, but for the entire world, and it is crucial that countries work together to establish global guidelines, given the transnational nature of technology and the digital economy.

This would not only enable a harmonization of AI practices, but also provide an equitable level of protection for the rights of citizens across the world. Certain that this could be a great opportunity for Brazil to take on a global role in a new regulation, which would bring more legal certainty and, with it, new business for the country.

However, regulating AI is a complex task that requires a delicate balance. On the one hand, it is essential to ensure privacy, security and fairness, on the other hand, regulations cannot stifle innovation, which is vital for economic growth and global progress.

Thus, as we move towards an increasingly AI-driven era, it is crucial to ensure that both this and other emerging technologies are used in a beneficial way, with careful regulations, ethical guidelines and a commitment to the well-being of all citizens, to shape the future of AI in a way that reflects our deepest values, with new regulations adjusted and adapted to the realities of countries, such as Brazil, imperative. And as the “terminator of the future” would say: “Hasta la vista, baby.”

https://www.estadao.com.br/politica/gestao-politica-e-sociedade/regulacao-de-ia-o-brasil-esta-pronto/