PL 2338/2023, commonly called the Marco Legal da IA, is Brazil's comprehensive AI bill. It was introduced in the Federal Senate on 3 May 2023, approved by the Senate on 10 December 2024, and is pending in the Chamber of Deputies as of April 2026. The proposed enacting body is the National Congress of Brazil. This page summarizes the bill as a source; the structured legislation page is Brazil AI Bill (PL 2338/2023) — Source Summary.
Origin and status
The bill was drafted by a 2022 Senate-appointed commission of jurists chaired by Justice Ricardo Villas Bôas Cueva. It tracks the EU AI Act (Regulation 2024/1689) risk-based structure while adding features described as distinctive to Brazil: individual rights building on the LGPD, a creator-remuneration provision for training data, and an explicit collective-rights dimension.
The Senate approved the bill on 10 December 2024. It was forwarded to the Chamber of Deputies on 17 March 2025, where a special committee was constituted on 29 April 2025 to handle review; the committee was established because the bill was referred to more than three standing committees. As of April 2026 the Chamber has not voted on the bill and no floor vote is scheduled.
Summary of provisions
The bill sets out a risk-based tiered framework with prohibited, high-risk, and general categories, with additional GPAI-style obligations added in the Senate amendments.
Prohibited AI systems include vulnerability exploitation; public-authority social scoring unrelated or disproportionate to behaviour; autonomous weapons targeting persons; and real-time remote biometric identification in public spaces, with narrow law-enforcement exceptions.
High-risk AI systems include those used in critical infrastructure, education, employment, essential public and private services, law enforcement and justice, migration and border control, and general-purpose AI with systemic impact. Obligations for these systems include algorithmic impact assessment, data governance, risk management, transparency, human oversight, and cybersecurity.
Individual rights under the bill cover information, explanation, contest, non-discrimination (with a presumption of discrimination on disparate impact), privacy, and participation in decisions affecting collective rights.
On copyright and training data, the bill provides a text-and-data-mining exception for non-commercial research only, an opt-out right for creators, and a remuneration arrangement for creators whose works are used for commercial AI training.
For governance, the bill creates a National System for Regulating AI (SIA); the Senate text designates the ANPD as competent authority, while sectoral regulators retain sectoral authority.
Penalties reach up to BRL 50M per infraction or 2% of group turnover in Brazil.
Key claims
Several claims about the bill are well supported. The Senate approved PL 2338/2023 on 10 December 2024. The bill was forwarded to the Chamber of Deputies on 17 March 2025, with a special committee established on 29 April 2025 because the bill was referred to more than three standing committees. The bill has not been voted on by the Chamber as of April 2026, and no floor vote is scheduled. The bill is structurally modelled on the EU AI Act (risk tiers plus GPAI) with additions described as distinctive to Brazil: creator remuneration, broader individual rights, and a collective-rights dimension.
Other points carry less certainty. Whether the ANPD will serve as the lead AI regulator is contested, with industry lobbying for a distinct AI regulator. The creator-remuneration provisions are reported as likely to face dilution pressure in the Chamber, with CISAC and Brazilian creators' organisations lobbying to preserve them. The real-time biometric identification prohibition is described as a point of contention given existing state-level deployments in São Paulo and Bahia.
Relation to other pages
The bill is structurally modelled on the EU AI Act (Regulation 2024/1689) while adapting rather than copying that template, a path of interest to other jurisdictions borrowing the EU structure. Its creator-remuneration provision connects to AI Copyright Litigation — Analysis and is described in CISAC's coverage as globally pioneering. The rights-rich approach contrasts with the pro-innovation orientation of the Japan AI Promotion Act — Source Summary. The bill is a parallel to the South Korea AI Basic Act — Source Summary, which pursued enactment in the same month (December 2024); Korea's effort succeeded while Brazil's remains pending. The bill also connects to AI Sovereignty through its assertion of Brazilian regulatory authority distinct from the EU, US, and China. It is the first Brazilian legislation covered here. No direct contradictions with existing pages were identified.
Relationships
- supports: Brazil AI Bill (legislation page)
- related: EU AI Act (Regulation 2024/1689), South Korea AI Basic Act — Source Summary, Japan AI Promotion Act — Source Summary, Canada AIDA (Bill C-27, Part 3) — Source Summary, Australia Voluntary AI Safety Standard — Source Summary
- related: AI Copyright Litigation — Analysis, AI Sovereignty, China — Interim Measures for the Management of Generative AI Services
Provenance
No PII, credentials, or sensitive identifiers appear in the source material.
Sources
- Senado Federal record (Matéria 157233):
https://www25.senado.leg.br/web/atividade/materias/-/materia/157233 - Library of Congress, "Brazil: Senate Advances Discussions on Bill to Regulate AI Use" (May 2025):
https://www.loc.gov/item/global-legal-monitor/2025-05-23/ - Lexology, "Brazilian Bill regulating the use of Artificial Intelligence is approved by Senate and goes to the Chamber of Deputies" (2025)
- White & Case, AI Watch: Global Regulatory Tracker — Brazil
- CISAC, "Creators Celebrate Brazil's Senate Approval of AI Bill" (Dec 2024)
- SIDI, "PL 2338/2023: the impacts of regulating Artificial Intelligence in Brazil"
- Adeptiv.ai, "Brazil Artificial Intelligence Act 2025"