- The Wheat from the Chaff
- Posts
- The Wheat from the Chaff #17
The Wheat from the Chaff #17
This week, we are going to talk about the regulatory framework of Artificial Intelligence (AI) in Brazil. But before that, did you know that we have already surpassed the milestone of 3,000 readers of our publication? I want to start today's text by expressing my gratitude for the trust of all of you! Let's get started!
Jhonata Emerick
August 17, 2023
As part of my job, every week I read a lot of content about data, technology, and artificial intelligence (also known as AI). I filter them and send you the best ones, every Sunday.
Hello everyone 👋 ,
This week, we are going to talk about the regulatory framework of Artificial Intelligence (AI) in Brazil. But before that, did you know that we have already surpassed the milestone of 3,000 readers of our publication? I want to start today's text by expressing my gratitude for the trust of all of you!
Let's get started!
Less is More
Today we will have a special edition, by popular demand, about the regulatory framework of Artificial Intelligence (AI) in Brazil. And a little curiosity: I had the opportunity to participate in some discussions on this topic at the Federal Senate!
Jhonata Emerick at the Federal Senate
Before we delve into regulation, it's worth noting that Brazil invests very little in Artificial Intelligence (AI) compared to other countries. For instance, in 2022, the United States invested approximately $48 billion in AI, about 3.5 times more than China, according to Stanford University. The data can be seen in the graph below.
Considering our low automation rate and existing inefficiencies in various sectors, as well as the size of our consumer market, Brazil can be seen as fertile ground for AI-based systems. Therefore, strategically crafted regulation can secure us a position in this complex puzzle.
In early May, the President of the Senate, Rodrigo Pacheco, introduced a bill aimed at establishing the legal framework for Artificial Intelligence in the country. The biggest challenge lies in finding a strategy that allows for reconciling the development and advancement of AI-based technologies with individual rights and guarantees.
The first bill focused on establishing principles for the use of Artificial Intelligence in Brazil emerged in 2019, aiming to ensure its development for the benefit of society. In 2020 and 2021, new projects emerged, addressing traceability of decisions, prevention against discriminatory biases, and governance issues. Based on these proposals, the Federal Senate created a committee in March 2022 to assist in the drafting of a preliminary bill aimed at regulating the development and application of AI in Brazil.
The final text, known as the Legal Framework for Artificial Intelligence, can be understood based on five pillars: Definition of Agents, Principles on Use and Development, Rights of Individuals, Risk Classification of AI Systems, and Governance Measures.
The text defines "AI Agents" as the providers who develop Artificial Intelligence systems and the operators who use them. According to the law, both are responsible for repairing any damages caused by these systems, regardless of the level of autonomy of the system.
The second pillar discusses important principles regarding the use and development of AI-based systems. These principles aim to ensure that the technology is employed responsibly. For example, AI should be designed and used in a way that does not discriminate against anyone. Additionally, it should be transparent, meaning that people should be able to understand its functioning and how decisions are made. These principles aim for the technology to be employed in an ethical and safe manner.
The third pillar focuses on individual rights, emphasizing that anyone who interacts with AI systems has clear rights regarding the nature of that interaction. If a system makes a decision that impacts someone, that individual has the right not only to demand an explanation but also to contest such a decision. Additionally, people have the right to request human intervention in certain decisions, as well as the right to privacy and protection of their personal data. These rights are guaranteed and can be exercised before the competent authorities.
AI systems are classified as 'excessive risk' and 'high risk,' according to the proposed law. 'Excessive risk' systems, which pose potential harm to health, safety, or individual rights, are not allowed in Brazil. On the other hand, 'high-risk' systems, although permitted, require heightened attention. This category includes systems employed in critical infrastructures such as energy and transportation, professional or credit assessments, autonomous vehicles, medical diagnoses and procedures, and biometric identification systems. The law stipulates that those responsible for 'high-risk' systems, whether providers or operators, must be held accountable for any damages caused.
The fifth component of the Legal Framework emphasizes the need for governance. Those who develop or use these systems must establish procedures and structures that protect individuals impacted by the decisions produced by their systems. This involves implementing transparency practices and data management that prevent biases and ensure privacy from the early stages of system creation.
The Legal Framework begins an essential discussion, establishing roles and responsibilities and creating a safer environment for everyone by preserving and guaranteeing fundamental rights, such as non-discrimination. The governance aspect initiates a process that ensures the necessary legal certainty for society to fully leverage the potential of this technology.
However, we need to consider that we still have limited knowledge about the technology to establish overly rigid concepts at this stage. Legislation should evolve in harmony with technology; otherwise, we run the risk of having an outdated law in a short period of time. Governance, while providing security, should be flexible enough not to become a barrier to the development of technology in appropriate ecosystems.
The bill will follow the constitutional process, but you might be wondering: How can I prepare myself? The first step is to catalog the Artificial Intelligence (AI) systems that your organization develops or employs. An initial assessment of these systems, including their types and applications, is crucial to start identifying potential risks. If you discover systems that can be categorized as high-risk, it will be necessary to begin formulating measures that may encompass management policies and impact assessments. Another essential step is to train professionals to work in the field of governance and auditing of AI systems, which present different levels of autonomy and complexity.
As we discussed in our edition #6, the topic of ethics has been debated by society since Aristotle, and we still have some long strides to take. Technological advancements in general accelerate debates and discussions, also driving legal evolution. The era of AI is just beginning! :)
Interesting topics in Brazil:
Trending topics Global:
We would love your help sharing this website!
If you know someone that might like our posts, don’t be shy and share it with them! By the way, we appreciate this!
See you Sunday…
We always arrive in your inbox around 12:08 PM. Some email servers are stubborn and delayed... Others are even worse and throw us into spam and/or promotions. What's the logic behind this timing? ...None :)