Establishes a framework of aims, foundation, and principles for AI governance in Brazil, defines guidelines for future regulations, including risk-proportionate regulation. Prioritising human rights, environmental protection, transparency, non-discrimination, competitiveness, and the balancing of innovation and self-regulation with safety requirements for high-risk systems.
Analysis summaries, actor details, and coverage mappings were LLM-classified and may contain errors.
This is a binding legislative instrument enacted by the Brazilian National Congress with mandatory language throughout, establishing legal obligations, enforcement mechanisms through sectoral bodies, and explicit liability provisions for non-compliance.
The document has minimal to good coverage of approximately 8-10 subdomains, with strongest focus on discrimination and bias (1.1, 1.3), privacy protection (2.1), misinformation risks (3.1), governance structures (6.5), and system robustness (7.3, 7.4). Coverage is concentrated in fairness, transparency, and governance domains, with limited attention to malicious actor risks or advanced AI safety concerns.
This is a horizontal framework law that applies across all economic sectors in Brazil. The document explicitly adopts a cross-sectoral approach with sectoral bodies responsible for implementation in their respective domains. While it mentions specific applications in education and public services, the law governs AI development and deployment across all industries operating in Brazil.
The document addresses multiple AI lifecycle stages with emphasis on deployment and operational monitoring. It covers planning through principles and foundations, data processing through privacy requirements, model development through security and bias mitigation requirements, validation through risk assessment requirements, deployment through transparency obligations, and operation through continuous monitoring and risk management provisions.
The document explicitly defines and covers AI systems broadly, including machine learning, knowledge-based systems, and statistical approaches. It does not specifically mention frontier AI, general purpose AI, foundation models, or compute thresholds. The focus is on AI systems generally without distinguishing between generative, predictive, task-specific, or open-weight models.
The document is a Federal Senate Bill enacted by the National Congress, as indicated by the header and signature. It was approved by the Chamber of Deputies on September 30, 2021.
The law establishes a distributed enforcement model where sectoral regulatory bodies with technical expertise are responsible for monitoring risk management, establishing rights and duties, and recognizing self-regulatory institutions. The Federal Executive Branch is responsible for regulations and strategic management.
The same sectoral bodies responsible for enforcement are explicitly tasked with monitoring risk management of AI systems in their areas of competence. The federal government is responsible for promoting strategic management and monitoring in the public sector.
The law applies broadly to all agents involved in AI development and operation in Brazil, including both private and public sector entities. It explicitly mentions agents in the development and operation chain, legal entities providing public services, and sectoral bodies with regulatory competence.
17 subdomains (4 Good, 13 Minimal)