CONFIRMED

Brazil's National Data Protection Authority (ANPD) revealed, at the Fale com a ANPD panel held on September 2, 2026 during the 17th Seminar on Data and Privacy Protection, that it is building its own artificial intelligence center inside the agency. A day earlier, on September 1, ANPD had already opened the window to receive public suggestions for its next regulatory agenda, focused on the 2027-2028 biennium. Both initiatives were reported by ConvergenciaDigital, Capital Digital, O Povo and Brasil em Folhas.

What ANPD's AI center is

The center is being built by the innovation lab of ANPD's Technological Innovation Superintendency. According to Lucas Anjos, the area's superintendent, the goal is to develop regulatory capacity both in infrastructure and in staff training and specialization, so the agency has real technical conditions to analyze AI systems within its remit, including automated decisions and applications that process large volumes of personal data.

The regulatory sandbox already in testing

Part of the same strategy is ANPD's AI regulatory sandbox. Its first cohort is in final testing and works on issues tied to Article 20 of Brazil's LGPD data protection law, covering transparency, observability and explainability of automated decisions. Three companies are participating, from medical diagnosis, transportation and financial sector security. The expectation presented at the seminar is that full results from this first stage will land by the end of October 2026, with a second sandbox edition expected to launch later this year and run through 2027.

Public consultation for the 2027-2028 agenda

A day before revealing the AI center, on September 1, 2026, ANPD opened the window for researchers, companies, civil society organizations, public institutions and citizens to submit contributions toward its next regulatory agenda. Suggestions can be sent until October 16 through the Brasil Participativo platform and will help shape ANPD's Regulatory Results Agenda, the agency's main planning instrument, which gathers priority themes, goals and deadlines for each regulatory project.

A movement building since December

Both initiatives continue a process that started in December 2025, when ANPD published its Priority Themes Map for the 2026-2027 biennium and included artificial intelligence and emerging technologies as one of the agency's four enforcement axes. Building an internal technical center and testing a regulatory sandbox are the first concrete steps toward turning that stated priority into real supervisory capacity.

Why it matters for Brazilian agencies and SMBs

Even without an approved AI legal framework, ANPD is already building, in practice, the structure that will oversee AI systems processing personal data in Brazil, based on the LGPD. Agencies and SMBs using AI for customer service, customer scoring or any kind of automated decision should watch the sandbox's results on transparency and explainability closely, since those criteria tend to become a practical compliance reference even before a dedicated AI law takes effect. Joining the public consultation before October 16 is also a direct way to shape those rules before they are finalized.

Sources

ConvergenciaDigital, ANPD terá novo sandbox regulatório de inteligência artificial e monta centro próprio de IA: https://convergenciadigital.com.br/governo/anpd-tera-novo-sandbox-regulatorio-de-inteligencia-artificial-e-monta-centro-proprio-de-ia/ | Capital Digital, ANPD terá centro de IA e reforça estrutura para fiscalizar novas tecnologias: https://capitaldigital.com.br/anpd-tera-centro-de-ia-e-reforca-estrutura-para-fiscalizar-novas-tecnologias/ | O Povo, ANPD recebe contribuições para agenda regulatória 2027-2028: https://www.opovo.com.br/noticias/brasil/2026/09/01/anpd-recebe-contribuicoes-para-agenda-regulatoria-2027-2028.html