The Temporary Commission on Artificial Intelligence of the Senate (CTIA) presented, on April 24th, the preliminary text of the substitute for the Artificial Intelligence Act in Brazil.
The document, prepared by Senator Eduardo Gomes, synthesizes ten other bills proposing guidelines for the use and development of AI in the country. Among them is Bill of Law no. 2.338/2023, authored by the President of the Senate, Rodrigo Pacheco, which aims to create the Legal Framework for AI.
The substitute adopts a risk-based approach and classifies AI systems into specific categories, such as “excessive risk” and “high risk.”, assessed through a preliminary self-assessment, which determines specific obligations and responsibilities for each category.
In the context of civil liability, the substitute bill envisages the obligation to compensate for damages, regardless of the system’s degree of autonomy, adopting interpretations of liability according to the risk presented by the AI system.
Copyright
There is also a section exclusively dedicated to copyright content, which includes provisions such as the obligation for suppliers to indicate the protected works that have been used in AI training processes and the possibility for authors to restrict the use of their works in AI systems.
The text also provides for the creation of a specific national authority, which will have normative, regulatory, and sanctioning powers and will be responsible for coordinating the National System of Regulation and Governance of Artificial Intelligence (SIA).
The proposal of the new text is still subject to changes based on the contributions from developers, jurists, and experts, who are expected to submit them by May 9th.
At this relevant moment for the national regulatory landscape, it is essential to have the active participation of the entire society, to promote a regulatory environment capable of ensuring the responsible development and use of AI in Brazil.
