Regulatory Landscape
Brazil is pursuing one of Latin America’s most ambitious AI regulatory initiatives, with Bill 2338/2023 advancing through the National Congress toward what would be the region’s first comprehensive AI law. The bill draws heavily on the EU AI Act’s risk-based approach while incorporating adaptations reflecting Brazilian social priorities, including strong provisions on algorithmic discrimination and digital inclusion.
Brazil’s regulatory ambition reflects both the scale of its AI adoption — as Latin America’s largest economy with a significant technology sector — and its tradition of comprehensive digital legislation, established through the Marco Civil da Internet (2014) and LGPD data protection law (2020).
AI Bill 2338/2023
The AI bill establishes a tiered system distinguishing between excessive risk (prohibited), high-risk, and other AI systems. High-risk categories include AI in public administration, criminal justice, healthcare, employment, credit scoring, and education. The bill creates rights for persons affected by AI systems, including the right to explanation, right to human review, right to non-discrimination, and right to correction.
Key provisions include mandatory algorithmic impact assessments for high-risk systems, requirements for human oversight, transparency obligations, and the establishment of a dedicated AI supervisory authority. The bill also addresses foundation models and generative AI, requiring providers to implement safety measures and maintain transparency about training data and system capabilities.
LGPD and AI
Brazil’s data protection law provides a foundation for AI governance through its provisions on automated decision-making. Article 20 of the LGPD grants individuals the right to request review of decisions made solely through automated processing that affects their interests. The ANPD (National Data Protection Authority) has issued guidance on applying these provisions to AI systems and is positioned to play a significant role in AI oversight.
Social Justice Emphasis
Brazil’s AI governance discussion is distinctively shaped by concerns about algorithmic discrimination in a deeply unequal society. Civil society organizations have documented how AI systems in criminal justice, credit scoring, and public services can perpetuate and amplify existing racial and socioeconomic inequalities. The AI bill’s strong provisions on non-discrimination and impact assessment reflect these concerns.
Industry Impact
Brazil’s AI sector spans fintech (Nubank, PagSeguro), agritech, healthtech, and enterprise applications, with significant adoption across banking, telecommunications, and e-commerce. Companies are monitoring the AI bill’s progress and preparing compliance programs, while international companies serving the Brazilian market anticipate obligations similar to those under the EU AI Act.
What Comes Next
Bill 2338/2023 is expected to advance through congressional debate in 2025, with the Senate and Chamber of Deputies potentially reaching agreement on final text. Key remaining debates involve the scope of the supervisory authority’s power, the balance between innovation promotion and risk management, and resource allocation for enforcement. Brazil’s approach will influence AI governance development across Latin America, where several countries look to Brazilian digital policy as a model.