On March 29, 2023, “The New York Times” published an article regarding an open letter signed by more than a thousand leaders in the technology sector, including Elon Musk, president of Tesla, Space X, and X (formerly Twitter). The letter requested a stop in the development of artificial intelligence above GPT-4, developed by OpenAI, considering the potential risk of threatening the very existence of society as we know it and even humanity.
There are Hollywood productions such as “The Terminator” and, currently, the movie “The Creator“, which address precisely the kind of threat that artificial intelligence brings to the human race, in a clash that is almost impossible to win.
The point is that artificial intelligence is no longer something only typical of science fiction movies, but it has become something that already impacts people's daily lives, sometimes without them knowing. This text itself could be written by an artificial intelligence tool, without people even suspecting that its creator was not human.
In the area of technology, artificial intelligence is increasingly becoming part of softwares and applications, whether in the form of “borrowing” already consolidated technologies such as the famous “GPT chat” or artificial intelligence technologies developed by the company that manufactures said softwares or applications. Programming languages such as Python, which is open source, favors the development of artificial intelligence.
This is an uncharted world, in which there are no restrictions, limits, rules, or even any regulation… at least in Brazil. This does not imply, however, that society and parliamentarians are totally unaware of this matter.
There are some non-governmental organizations paying full attention to the topic, such as, for instance, Coalizão Direitos na Rede, which has already published an open letter for the regulation of artificial intelligence in the country, highlighting some very relevant points, such as:
1. The fear that artificial intelligence could reinforce or create forms of discrimination based on race, gender, sexual orientation, social class, and ideology.
2. The fact that there is not only a focus on data management, but also on solutions to issues such as algorithmic opacity, discrimination, and the delimitation of risk levels for different artificial intelligence applications.
3. The creation of thematic groups that guarantee the participation of experts from other areas of knowledge and who have a deep understanding of technical aspects and social impacts involving artificial intelligence systems.
4. The convening of public hearings to allow multisectoral, interdisciplinary, and representative participation of the diverse groups affected by this issue.
From a regulatory perspective, there are currently 3 bills under the Brazilian National Congress on the topic:
Out of the 3 bills described above, the first and third basically list general principles that should be observed in the development of artificial intelligence. Unfortunately, their content is dramatically superficial in defining the limits of artificial intelligence, in order to prevent malicious development or even enable self-management, which could threaten humans.
The second bill is much more robust, starting to address important issues that should be on the authorities' radar. This bill is currently awaiting a public hearing for public contributions to its original text.
In Brazil, the Brazilian Data Protection Authority, for instance, has already started to address this topic in a document called “Sandbox Regulatório”, making it clear in a very lucid way how artificial intelligence can impact the practical experimentation of rules in protected environments, in order to test their effectiveness.
In fact, this topic is so complex and deserves so much attention from the authorities that it will not be a surprise if an agency is created with the specific purpose of regulating and monitoring the creation and use of artificial intelligence from now on.