The Brazilian legislative landscape is making significant strides in understanding and regulating Artificial Intelligence, recognizing its potential impacts on society. Here, Nathalia Santos, a specialist in technology and intellectual property at Pellon Advogados, shares an update on the key points.
Recently, on April 24th, the legislative branch released a preliminary report regarding the regulation of artificial intelligence in Brazil, introducing the concept of generative Artificial Intelligence in an unprecedented manner. The aim of this new report is to integrate a series of other ongoing legislative projects, scrutinized by a temporary technical commission dedicated to the topic.
In addition to addressing the innovation of generative AI, the report highlights rights protection and the establishment of the National System for Regulation and Governance of Artificial Intelligence (SIA), aiming to create a coordinated regulatory environment under an authority designated by the Executive branch.
Although preliminary, the report outlines important concepts and already provides a clear direction on the principles that will be addressed in the regulation of artificial intelligence usage in Brazil. Key points such as modulation of regulation in risky activities, civil liability, cultural and artistic rights, intellectual property, as well as the protection of individual, social, and economic fundamental rights are directly addressed and will undoubtedly be covered in a final regulation.
The new report is not merely a compilation of various legislative projects circulating in Congress. It also presents some novel points or enhanced versions of previous proposals. Several crucial points stand out for future discussions:
- The emerging regulation is heavily inspired by the European model, somewhat more assertive in principles and regulation, with a heavier modulation of regulation in risky activities, especially in profiling activities where a machine/robot selects characteristics from a database to separate (and judge) humans.
- Nonetheless, there are also points inspired by the model applied in the United States, one of which is the creation of the SIA, whose role is to be an articulating authority within the already regulated ecosystem, unlike that seen in the European Union.
- As part of a growing global concern, autonomous weapons (those that, when activated, can select and attack targets without additional human intervention) are prohibited from use “without significant human control, whose effects are unpredictable or indiscriminate, or whose use implies violations of International Humanitarian Law.”
- Principle-based issues will be crucial in aiding interpretations in future analyses and judgments when concrete cases are in dispute in the Judiciary and administrative bodies
- Nevertheless, civil liability and consumer rights protection have a well-advanced definition, and it is expected that such situations will be judged in accordance with existing legislation.
- One of the main issues on the agenda of the new regulation, copyright and intellectual property protection, is addressed, foreseeing the obligation of platforms to use databases with protected materials; disclose which works were used, paving the way for charging fees by authors and rights holders, which will create a new market and standards for content producers and owners, also fostering a market for synthetic or artificial content created solely for machine training.
- The text also recognizes the self-regulatory capacity of specific sectors and the prominence of codes of conduct and policies to self-limit the capacities of each economic sector, although it does not define the issues of liability between AI platform operators and their creators, leaving it to the evaluation of judges, administratively or in the Judiciary, the accountability in case of failures or problems, according to each specific case, following Brazilian civil liability principles.
Without neglecting the valid efforts and advancements in crafting a legal framework that brings greater legal and social security in the use of AI tools, the possible adoption of a more stringent European model raises questions about the impact on experimentation and innovation in Brazil, especially due to the absence of provisions for regulatory tests.
