In recent months, bills to regulate the use of artificial intelligence (AI) technology in the country have been advanced in the legislature. The most advanced proposal from the chamber, written by Representative Eduardo Bismarck (PDT-CE), is ready for a vote in the House plenary. Experts considered the projects to have positive points, but said that regulation may be premature, given the speed with which AI technology is developing.
In fiction, AI is often portrayed in menacing stories, sometimes involving machines rebelling against humans. She is, for example, in films such as 2001: A Space Odyssey (1968), or The Matrix (1999). In real life, artificial intelligence is a type of computer program that is able to interpret data, learn from it and make decisions independently to accomplish a particular task set by its creator.
Today, artificial intelligence exists in a series of everyday actions. Algorithms are also found in online stores, in company inventory monitoring, in facial recognition tools, fraud prevention systems, and in analyzing consumer behavior patterns.
In the room, there are at least three other projects in addition to the one presented by Eduardo Bismarck: the proposals of Representatives Leo Moraes (Podemos-Ru), Bosco Costa (PL-SE) and Gustavo Frue (PDT-PR). They are all processed together, and attached to the Bismarck Project. In the Senate, AI is subject to three more bills. The most advanced was written by Veneziano Senator Vital do Rigo (MDB-PB) and has as rapporteur the head of government in the Senate, Eduardo Gómez (MDB-TO).
The initial version of Project Bismarck, which is being processed in the room, is simple: there are only nine articles, outlining the general principles that should govern the use of AI algorithms in Brazil. According to the proposal, this type of program should be built with respect for principles such as human dignity, protection of personal data, non-discrimination, transparency and security. Rapporteur of the proposal is Representative Louisa Kenziani (PTB-PR).
The room script also creates an “artificial intelligence agent” character, who can be either the developer or operator of the program. An “artificial intelligence agent” is legally responsible for the decisions made by the algorithm. The agent is also responsible for ensuring that the software complies with the rules of the General Data Protection Regulation (LGPD).
The Senate bill is more synthetic, with six articles. Similar to the Chamber’s proposal, the Venezuela text states that AI development in Brazil respects principles such as ethics, human rights, democratic values, and protection of privacy, among others. The project entered the agenda and received nearly 20 amendments, but was withdrawn so that the rapporteur could improve the final text. One amendment by Senator Paulo Baim (PT-RS), to specify that AI adoption takes into account the impact on jobs, including in the public sector.
In recent months, the House and Senate have held hearings to discuss potential regulation of artificial intelligence. Class entities representing companies in the technology sector demonstrated against the regulations. The fear is that the new rules may restrict the development of the technology, whose potential and future implications are still unknown.
Bismarck said he realized the need for the project when he saw other countries making progress on the topic, by creating laws on technology, often based on principles outlined by the Organization for Economic Co-operation and Development (OECD). “We have established principles, rights and duties, to be able to help technology evolve. We are not getting into little bottlenecks. This will be left to complementary legislation later, if necessary, when the technology is more advanced,” he said.
The deputy also rejects the idea that the project could hinder the development of the technology. He declared, “Today’s big corporations no longer want to deal with borders and rules, because they believe their compliance goes beyond local laws. But they are not sovereign nations capable of laying down these principles, as Parliament can.” “Our proposal is in line with the law of the United Kingdom, Singapore, Japan and the United States. Why is there no such criticism?” asked the congressman.
Technology under discussion
The room has at least four projects in the field of artificial intelligence. In addition to the proposal of Representative Eduardo Bismarck (PDT-CE), which is the most advanced, Leo Moraes (Podemos-Ru), Bosco Costa (PL-SE) and Gustavo Frue (PDT-PR) also made proposals on this topic. They all go together.
The regulation of artificial intelligence technology has also been included in three bills submitted to the Senate. What is more advanced, to date, is written by Veneziano Vital do Rêgo (MDB-PB) and has as Rapporteur the Head of Government in the Senate, Eduardo Gomez (MDB-TO). In short, the text contains only six articles.
Veneziano Vital do Rêgo’s proposal specifies that the development of AI in Brazil respects principles such as ethics, human rights, democratic values, and protection of privacy. Similarly, the Bismarck Project states that software should be built with respect for principles such as human dignity, transparency and security.
“Friendly zombie guru. Avid pop culture scholar. Freelance travel geek. Wannabe troublemaker. Coffee specialist.”