Brazil’s AI Legal Framework Still Awaits a Chamber Vote

Brazil’s proposed Marco Legal da Inteligência Artificial—PL 2338/2023—passed the Senate unanimously in late 2024 but still awaits a rapporteur’s report and floor vote in a Chamber of Deputies special commission. In April 2026 the Chamber attached the Executive’s PL 6237/2025, which would create a National System for AI Development, Regulation and Governance (SIA), so the two files now move together under priority handling.

Filed under Policy and dated September 10, 2026, this AI4Brazil briefing stresses that nothing in those bills is binding law today. Meanwhile the national data-protection authority (ANPD) has continued practical experimentation, including sandbox-style monitoring that published partial results in mid-2026. Startups and multinationals therefore face a familiar Brazilian pattern: political delay paired with administrative learning.

Why it matters: São Paulo product teams, Brasília compliance leads, and investors need to know whether risk-tiered transparency duties and a residual ANPD role are coming. Uncertainty has a cost—governance designs get postponed, and foreign capital stays cautious. ABStartups-linked commentary has warned that over-heavy compliance could chill the roughly fourteen hundred AI-core startups in the ecosystem.

What it means in practice

Brazil’s AI Legal Framework Still Awaits a Chamber Vote — contextual photo

Do not wait for the gavel to inventory systems. Map which AI tools affect credit, hiring, health, or essential services; document training data sources and human override paths; and run time-boxed pilots with exportable logs. Prefer vendors who can support eventual risk assessments without rewriting your stack. Treat Chamber schedules as planning signals, not deadlines.

Workers and consumers deserve clarity when automated decisions touch rights. Brazilian firms that retain trust usually disclose limits early. AI4Brazil will keep underscoring that point.

What to watch next: the rapporteur’s opinion; how SIA coordinates with sector regulators; and ANPD sandbox lessons that may foreshadow obligations. Readers can continue on the AI4Brazil homepage for related stories, or browse the Newsroom for additional briefings.

Bottom line: treat this update as orientation, not instruction. AI rulemaking in Brazil is real, uneven, and unfinished. Organizations that benefit most will prepare documentation now, measure pilots honestly, and keep people accountable for outcomes.