Brazil, Chile advance AI laws

Brazil's Science, Technology and Innovation Committee in the Chamber of Deputies approved a substitute text on the use of data by artificial intelligence, based on PL 4089/2024, 5441/2025 and 6586/2025. The proposal remains under conclusive review, and still has to pass through the Consumer Protection Committee and the Constitution and Justice and Citizenship Committee, then the Senate, and, if it moves forward, presidential approval.

What changed in the approved text in Brazil?

The approved opinion in Brazil corrects the original approach and makes it explicit that data protection and processing rules also apply to data used by artificial intelligence systems, through changes to the Marco Civil da Internet. It also adds transparency obligations, human review, and a prominent notice when users interact with content produced by AI.

According to Bom Dia Sorocaba, based on a note from Agência Câmara, the substitute text drops the idea of directly amending the LGPD and instead relies on the Marco Civil da Internet to frame these obligations. The same material says service providers will have to clearly disclose when users are interacting with content produced by artificial intelligence systems.

What is happening with PL 3.630/2025 in Brazil?

PL 3.630/2025 has already been approved by the Chamber of Deputies, moved to the Senate, received a favorable opinion from the Public Security Committee, and is now under review in the CCJ with a favorable report from Senator Esperidião Amin. The bill, backed by federal deputy Bia Kicis, has drawn criticism over its impact on personal data and due process protections.

Idec, within the Working Group on Surveillance, Platformization and Data of the Direitos na Rede coalition, sent a letter to the Senate and argued that the proposal could weaken the LGPD and undermine the presumption of innocence. According to Convergência Digital, the objection is that the text would allow the processing and disclosure of images, audio, and other personal data captured in alleged flagrante delicto situations inside commercial establishments.

The same network of organizations says expanding the grounds for processing and sharing data for security purposes threatens human dignity, the protection of personal data, the presumption of innocence, and due process, and is asking the Senate to review or shelve the proposal.

How is AI regulation advancing in Chile?

The Chilean bill regulating artificial intelligence systems, Bulletin 16821-19, consolidated with 15869-19, remains in its second constitutional stage in the Senate, assigned to the Committee on Future Challenges, Science, Technology and Innovation. The text was introduced by the executive branch in May 2024 and approved in detail by the Chamber of Deputies on October 13, 2025.

According to ArtecLaw's analysis, the Senate committee is still hearing technical testimony and there is no floor vote date yet. The same report says the initiative includes algorithmic governance obligations for providers, including impact assessments for high-risk systems, transparency requirements on the data used to train models, and stronger supervision duties when personal or sensitive data is processed.

ArtecLaw also notes that these requirements are set to align with Chile's future personal data protection law. In parallel, the Chilean Chamber of Deputies' website maintains a bill-tracking system that allows users to check the assigned committee, discussion stage, and sponsors of initiatives affecting digital policy and data protection.