This Q&A is part of the Brazil Program’s ongoing efforts to examine the evolving digital policy landscape in Brazil, particularly in the context of the 2026 electoral cycle and broader debates on artificial intelligence, platform governance, and digital infrastructure.
As countries around the world grapple with how to regulate digital technologies, Brazil presents a distinctive case. Recent developments have placed courts and regulatory authorities at the center of rulemaking, even as Congress continues to debate comprehensive frameworks. At the same time, the electoral calendar is reshaping the pace and feasibility of advancing new legislation.
In this Q&A, we speak with Brazil Program senior fellow Natalia Paiva to explore how Brazil is navigating key regulatory choices on artificial intelligence, platform accountability, and digital markets, as well as how the country is positioning itself within the global digital governance landscape.
How do you assess how Brazil’s electoral authority moved to tighten accountability around AI use in the 2026 elections? What are the implications for the broader digital regulatory agenda? How does Brazil differ from other countries?
Natalia Paiva (NP): As debates continue over bills regulating digital platforms and artificial intelligence, as well as their enforcement, Brazil enters 2026 with a more regulated digital environment. In the absence of a specific congressional law, the TSE (Electoral Supreme Court) has taken a central regulatory role—enforcing Resolution 23.732/2024, which bans deepfakes, mandates labeling of AI-generated content, and restricts political bots.
Electoral integrity is becoming a pillar of Brazil’s broader digital governance agenda. The Supreme Court’s June 2025 ruling declared Article 19 of Brazil’s Internet Civil Framework (Marco Civil da Internet) partially and progressively unconstitutional, strengthening platform accountability across all contexts. Brazil’s Digital Child Protection Framework (ECA Digital), set to enter into force in March 2026, adds diligence obligations that will also shape how disinformation and platform responsibility rules are interpreted in elections.
What makes Brazil distinctive is that this accountability push is driven largely by Electoral Justice itself, rather than by comprehensive AI legislation. Brazil’s model is judicially led: courts and regulatory authorities are shaping the rules while Congress is still debating a national AI framework.
Brazil is entering an election cycle, which typically shortens the legislative window for complex regulatory debates. Recently, the expiration of the measure that created incentives for data centers (REDATA) has drawn attention to what may lie ahead for other regulatory initiatives, including the legal framework for AI.
In your view, what does this episode signal about the political feasibility of advancing tech regulation in 2026?
NP: The expiration of the measure that created incentives for data centers (REDATA is a provisional measure that expired without a Senate vote) sends a clear signal about how tech regulation will move in 2026: provisional measures are a fragile instrument for complex agendas, but the government’s rapid response, introducing a replacement bill with urgency approved, shows that digital infrastructure and AI remain strategic priorities even under political constraints. The lesson is not that the regulatory agenda is stalled, but that the instrument’s form matters as much as the political will behind it.
With the electoral calendar narrowing the legislative window from the second semester onward, 2026 will be a year of selective progress. Three conditions determine what advances: existing legislative momentum, clear economic rationale, and cross-party support. Initiatives that check all three will move. Those that require deep technical consensus or involve contested institutional design—like REDATA itself or digital markets regulation—will move more gradually.
With that filter in mind, the initiatives with the most realistic chances are: the new data center bill replacing REDATA, which already has urgency approved; Bill 2338/2023—the AI framework approved by the Senate in December 2024 and now in the Chamber, which the government has declared a priority; platform work regulation (Bill 152/2025), flagged as a first-semester vote with unusual cross-party support; and the streaming bill (Bill 2331/2022), expected to resume in the first semester. More structural agendas, such as digital markets regulation and the cybersecurity framework, will likely see incremental progress at best before the electoral calendar takes over.
How do you assess the current legislative proposals for AI, including Bill 2338/2023, and what are the key institutional choices Brazil will need to make to balance innovation, accountability, and regulatory capacity?
NP: Brazil is entering a decisive phase. The main legislative initiative—Bill 2338/2023, approved by the Senate in December 2024 and now in the Chamber, proposes a risk-based framework with the National Data Protection Authority (ANPD) at the center of governance.
The key debate is not whether to regulate AI, but how to structure that regulation. A model centered on an independent ANPD offers predictability and coherence with Brazil’s data adequacy agreement with the European Union (EU). A model with stronger executive participation may better coordinate AI policy with industrial and sovereignty goals, but it introduces political risk. The government’s complementary proposal seeks to formalize that participation while preserving the ANPD’s regulatory role.
Beyond institutional design, two challenges stand out.
First, intellectual property: how to handle the use of copyrighted data to train AI models, and whether rights holders should be compensated—a debate that is both technically and politically complex. This has been a top concern for the federal government, but I think it’s a misplaced priority. Copyright compensation for training data is a legitimate debate, but it doesn’t belong in an AI framework bill. It risks overloading the legislation and distracting from what should be Brazil’s real focus: economic development and inclusion, ensuring that Brazilian companies and workers are well-positioned for the transformations ahead.
Second, regulatory capacity: as the ANPD takes on more responsibilities beyond strict data protection—AI, ECA Digital, data protection—its technical and operational capacity needs to keep pace with its expanding mandate, in terms of expertise, budget, and headcount.
Globally, we are seeing different regulatory models emerging, from the EU’s ex ante approach to digital markets to the more enforcement-driven model in the United States. As Brazil debates new rules for digital markets and platform governance, how do you see the country positioning itself within this global regulatory landscape?
NP: Digital regulation is evolving along different models. The EU has adopted a preventive, ex-ante approach with structural obligations for large platforms—though it is now recalibrating through the Digital Omnibus agenda, which seeks to reduce complexity and overlapping obligations. The US continues to rely on case-by-case enforcement with no federal AI or platform law. These are genuinely different bets on how to govern the digital economy, based on different worldviews and geopolitical agendas.
Brazil is not just watching from the sidelines. In January 2026, Brazil and the EU formally recognized the mutual adequacy of their data protection regimes—placing Brazil in a small group of jurisdictions with facilitated access to the European market. In February 2026, Brazil played an active role at the India AI Impact Summit, reinforcing its position as a relevant actor in Global South AI governance. These are concrete signals of where Brazil is anchoring itself internationally.
Domestically, the picture is mixed. The ECA Digital has been a lesson to the world in well-balanced child protection legislation. But Bill 4675/2025 proposes a digital markets framework inspired by the EU’s Digital Markets Act—at a moment when the EU itself is already revisiting its Digital Markets Act (DMA) through the Omnibus package. In this particular case, Brazil’s likely path is a hybrid one: combining preventive regulation with existing competition tools, while aligning digital governance with broader goals of innovation, competitiveness, and technological sovereignty. How that balance is calibrated will define Brazil’s role in the global digital economy.
Get it right, and Brazil becomes a reference for how middle-income democracies can govern digital markets, without importing models from Brussels or Washington, and without reaching for simplistic solutions to complex problems. That requires understanding our own societal and economic challenges and how digital transformation can address them. Only then can Brazil find a regulatory path that is nuanced, balanced, and genuinely its own.
