Japan’s House of Representatives approved, at the end of April, the first bill regulating Artificial Intelligence in the country. AI regulation is a topic driving discussions in legislatures worldwide, with different strategies in various countries. According to UNESCO, since 2016, over thirty nations have created laws focused on AI.
“In the case of the Japanese bill, the text does not foresee penalties, in order not to hinder technological innovation and to prioritize attracting investments to the field. Thus, to mitigate the risks of the technology to the population, existing laws and regulations will be used, such as the Penal Code,” explains Dr. Patrícia Peck, an expert in Digital Law and CEO of Peck Advogados.
The lawyer also summarizes the legislation on the topic that is in effect and makes comparisons. “The European Union, in 2024, approved the AI Act, a more comprehensive regulation, which is coming into effect in stages. At the beginning of the year, prohibitions on certain AI systems and requirements related to AI literacy came into force. In August, parts related to five other areas will come into effect, among which governance and penalties stand out. Since then, we can already observe major companies investing in team education.”
Regarding the US and China, the lawyer adds: “Despite not having a federal law so far, the US already has sectoral regulations and White House guidelines. China, on the other hand, has a highly regulatory model, focusing on national security, regulation of recommendation algorithms, and the use of generative AI.”
Regarding the European Union’s regulatory model, Patrícia Peck highlights the intention to provide robust protection. “The text creates strict rules for sensitive uses of AI, which prevents excesses in less threatening applications. In cases of high risk, the law provides obligations of transparency, explainability, and independent audits as central requirements. Additionally, the text also includes a mechanism for periodic review, which can keep the law updated over time,” she details.
Regulation in Brazil
Regarding Brazil, the expert cites Bill 2338/2023, which is under consideration in the Senate and aims to create a legal framework for the use of AI. “The bill regulating Artificial Intelligence faces challenges, such as the fact that large technology companies may influence the process, the possibility of the legislation delaying the development of new technologies and becoming obsolete quickly. However, it is based on principles such as respect for Human Rights, non-discrimination, transparency and security, accountability, and responsibility,” she details.
According to the lawyer, the approach to AI regulation in Brazil is risk-based, similar to the European Union’s model. “The bill proposes stricter rules for high-risk AI applications, such as facial recognition and automated judicial decisions, which is a positive point for the population,” she explains.
Highlighting other positive aspects, the expert emphasizes public participation in drafting the bill—which included experts, academia, and the private sector—and the encouragement of innovation and technology development in the country. “The involvement of regulatory bodies is necessary so they can contribute with a sectoral view of technology use,” she notes.
However, Patrícia Peck explains that the text could be more precise. “The text is excessively general; some critics point out that the law is too vague on crucial points, which may hinder its application. Additionally, the framework may quickly become outdated, given the speed of technological advancements,” she details.
Finally, if approved, the text may also spark other debates. “There are doubts about the structure and resources of the institutions that will be responsible for overseeing and enforcing the law,” the lawyer emphasizes.
