The Brazilian government’s executive branch hosted a political ceremony last week in which it unveiled its “Digital Brazil Agenda,” which proposes six government projects to build a “safer, more competitive, and more innovative digital environment.” The most high-profile of these was the Digital Child and Teenager Act, which would set rules for how social-media platforms must manage use of their products by minors. But just as important was Bill 4.675/2025, the long-awaited ex-ante digital markets regulation bill, which was formally submitted by President Luiz Inácio Lula da Silva’s government to the Brazilian House of Representatives.
The submission marks the culmination of a public consultation that the Ministry of Finance kicked off in January 2024 (see here and here). Indeed, the Brazilian antitrust community had been eagerly awaiting the government’s next steps ever since the ministry published its report on the matter last October. And while Bill 4.675/2025’s fate remains uncertain, it appears to have much stronger prospects of passage than its predecessor, which languished amid a political stalemate.
In this post, we will explore four key aspects of the bill. To start, we will break down its substantive features, including the criteria for designating companies as having “systematic relevance in digital markets,” the menu of possible special obligations, the absence of efficiency or justification defenses, and the proposed sanctions.
Next, we explain why it matters that the legislation is proposed as an amendment to the Brazilian Competition Law (BCL), rather than a freestanding regulatory regime, as well as how this affects its objectives and the limits to intervention. We argue that Bill 4.675/2025 could clash with existing competition law in various ways.
Third, we examine the proposal’s institutional design and, in particular, the decision to create a new Digital Markets Superintendency. There are obvious potential risks that could arise from having two investigative bodies within the agency. We argue that this could exacerbate legal fragmentation and the proliferation of conflicting standards and goals.
Finally, we offer reflections on the Brazilian political and economic context, the bill’s prospects in Congress, and the importance of conducting regulatory-impact assessments before moving forward. Unfortunately, there are currently no plans for any such assessments.
What Would Bill 4.675/2025 Do?
Bill 4.675/2025 proposes a form of ex-ante competition regulation for digital markets that in many ways mimics the features of the European Union’s Digital Markets Act (DMA), the United Kingdom’s Digital Markets, Competition and Consumers Act (DMCC), and Germany’s Section 19a GWB. Like those initiatives, it is an asymmetric regime designed to apply only to large technology platforms. According to government estimates, between five and 10 companies would meet the criteria for designation and thus become subject to the bill’s “special obligations.”
Systematic Relevance in Digital Markets
Rather than use the EU’s “gatekeeper” designation, the bill would instead target firms with “systematic relevance in digital markets,” more aligned with the DMCC’s “strategic market status.” Under Article 47-C, a company could be deemed systematically relevant when, among other factors, it has:
- presence in one or more multi-sided markets;
- market power associated with network effects;
- vertically integrated operations in adjacent markets;
- a strategic position that enables third-party business activities;
- access to a significant quantity of relevant personal or commercial data;
- a significant number of business users or end users; or
- a broad portfolio of digital products or services.
In other words, these are the structural and behavioral characteristics that the authors suggest renders a firm capable of generating competition risks that justify ex-ante intervention.
Designation Procedure
Article 47-C, § 1º sets minimum revenue thresholds: a company must have either at least 5 billion Brazilian real (roughly $940 million) in annual revenue in Brazil or R$50 billion globally (roughly $9.4 billion) to be designated as having “systematic relevance in digital markets.” This test would be applied to the entire corporate group, including both the corporate parent and any subsidiaries. The newly established Digital Markets Superintendency would be empowered to open the designation process, but the Administrative Council for Economic Defense’s (CADE) Administrative Tribunal would make all final determinations. Designations would expire after 10 years, when a new procedure is required for renewal (Art. 87-A, § 1º).
Special Obligations
Once a firm is designated, CADE launches a second administrative process to determine the obligations to which the company will be subject. Unlike the DMA, the bill does not impose a pre-set list of “do’s and dont’s.” Rather, these individualized assessments make it, in a sense, a “more flexible” and case-specific approach, similar to the UK’s DMCC model. Article 47-E provides a nonexhaustive menu of obligations that CADE can impose, including:
- submitting all M&A transactions to CADE for review, regardless of turnover thresholds;
- disclosing clear terms of service, ranking criteria, and price structures to users and business partners;
- refraining from self-preferencing, tying, foreclosure practices, discrimination, and predatory strategies;
- offering data portability, interoperability, and access to performance-measurement tools;
- allowing the installation/uninstallation of apps, enabling default changes, and providing adaptation periods when terms change; and
- offering products and services on nondiscriminatory terms and creating dispute-resolution mechanisms.
CADE can tailor these obligations to specific services and implement them via technical redesign or contractual changes, giving the council substantial discretion to calibrate remedies.
Defenses and Justifications
The bill’s primary substantive weakness is that it lacks any efficiency defense. Article 87-B, § 1º requires CADE to provide an “economic justification” before imposing obligations (a welcome procedural safeguard, and one that the DMA lacks) but it would not explicitly grant companies the right to argue that their conduct generates efficiencies that outweigh potential harms. This is at odds with Brazil’s existing antitrust framework, which recognizes efficiency defenses in both merger-review and unilateral-conduct cases (Art. 88, § 6º, I, c, and Art. 36, § 1º of BCL).
Internationally, the EU DMA is silent on efficiencies, while the UK DMCC explicitly allows a “countervailing benefits exemption.” By omitting such a defense, Bill 4.675/2025 risks overdeterrence that could suppress procompetitive innovation. It also risks undermining the integrity of Brazilian competition law by scrutinizing conduct under a different standard. Adding a statutory efficiency defense would bring the law closer to CADE’s established analytical tradition and to best practices.
Sanctions
Art. 47-G dictates that a designated firm’s noncompliance with the special obligations imposed by CADE would trigger penalties that mirror the existing sanctions in Brazilian competition law: fines that range from 0.1% to 20% of the company’s revenue in the “affected line of business,” as well as daily penalties for ongoing violations. The BCL does not offer precise guidelines on how to calculate the fines, which have developed via years of CADE’s precedents, and CADE would retain some discretion for these new cases, as well.
Participatory Mechanisms and Inter-Agency Cooperation
The bill requires public hearings and comment periods before CADE finalizes designations or obligations, adding a higher degree of transparency and public input into the process (Art. 48, § 2º). Article 47-F of the bill also calls for cooperation with other sectoral regulators (e.g., Brazil’s telecom and data-protection agencies) to ensure technical alignment. While this could be positive, without a clear allocation of responsibilities, the provision as written is both broad and vague in ways that risk jurisdictional conflicts between the agencies. Memoranda of understanding and formal cooperation mechanisms will be needed to boost legal certainty.
An Amendment, Not a New Regime
If enacted, Bill 4.675/2025 would bestow CADE with new powers and add various regulatory features, but the rules would remain firmly within the boundaries of Brazilian competition law. This differentiates it from the DMA, which is technically distinct from EU competition law (here), and makes it more similar to Germany’s Section 19(a).
This choice could have major implications for the bill’s future implementation and enforcement. As a subset of the Brazilian antitrust regime, Bill 4.675 would have to follow the same objectives and guiding principles. Historically, CADE has generally interpreted BCL’s goal to be consumer welfare: lower prices, higher output, improved quality, and enhanced innovation.
The Bill’s Stated Goals and the Need for Limiting Principles
Article 47-B articulates three objectives for the bill’s ex-ante digital provisions: “I – reduction of barriers to entry; II – protection of the competitive process; and III – promotion of freedom of choice.” The reference to “protection of the competitive process” is particularly vague and ambiguous. As Nicolas Petit and Lazar Radic have written:
The “protection of the competitive process” is similarly redundant. All acts of bad conduct spelled out in antitrust statute epitomize “distortions of the competitive process”. Collusion removes independence from competitors and monopolization eliminates rivalry. Both standards add nothing to text law and are thus circular, with Herbert Hovenkamp calling them “slogans”.
“The competitive process” is rudderless without both a measurable goal and a yardstick: anything can be said to be part of, or an exception to, the “competitive process.” This is likely why modern proponents of the “competitive process” inevitably end up incorporating elements of the consumer-welfare standard. For example, in his proposed framework of analysis, Tim Wu refers to the “raising of rivals’ costs” and “anticompetitive effects,” both terms of art that are central to the consumer-welfare paradigm of antitrust law.
Further, it is not immediately clear that this language maps neatly onto the consumer-welfare standard that CADE typically applies in its decisions. For instance, Article 36, § 1º of the BCL sets out an important limiting principle when it clarifies that mere size or market power is not unlawful when it is obtained on the merits:
[A]chieving dominance in a market by natural process and by being the most efficient economic agent in relation to competitors does not characterize the violation set forth in item II (“II – to control the relevant market of goods and services”) of the head provision of this article.
By contrast, when read in isolation, Bill 4.675/2025 appears to lack explicit constraints to ensure that ex-ante interventions actually benefit consumers, rather than merely protecting less efficient rivals or punishing size. For example, Article 47-C of the bill would treat “vertical integrations and operations in adjacent markets” as a characteristic that justifies designating a company as having “systematic relevance in digital markets.” While vertical integration can sometimes raise and cause foreclosure concerns, antitrust economics has long recognized that it typically generates substantial procompetitive benefits—such as eliminating double marginalization, reducing transaction costs, and facilitating investment in complementary products. Thus, treating vertical integration as an almost per-se risk factor would turn a ubiquitous and often efficiency-enhancing business strategy into a basis for enhanced regulatory scrutiny.
A similar concern applies, for instance, to the bill’s condemnation of self-preferencing conduct listed in Article 47-E, IV, c. While some forms of self-preferencing may harm competition and consumers, others can be neutral or procompetitive. Pre-installing or favoring a platform’s own apps, software, or websites (maps, payments, browsers) may improve a company’s product or the user experience, or ensure data security. Without a clear efficiency defense or CADE’s requirement to demonstrate harm to consumer welfare, such practices could be prohibited even when they lower prices, enhance innovation, or improve service quality.
This is at odds with CADE’s practice under BCL, where it has never treated self-preferencing as per-se harmful. As Dario Oliveira Neto has written, while self-preferencing was not thoroughly analyzed by CADE, the agency tends to view self-preferencing as potentially procompetitive and far from inherently anticompetitive. And as a previous study found, “when we look at markets in general, there have been few convictions for self-preferencing in CADE’s jurisprudence, with a conviction rate of 27% over the past 10 years.”
In this light, Bill 4.675/2025 risks working at cross-purposes with Brazilian competition law. That should not be surprising: the DMA, the “patient zero” of digital ex-ante regulation, was never designed with consumer welfare in mind. In fact, it explicitly disregards economic efficiency and consumer welfare as relevant benchmarks. Principles and presumptions drawn from the DMA thus sit uneasily within an efficiency-based, consumer-welfare–oriented antitrust framework.
If the government insists on introducing DMA-like provisions to Brazil, the BCL should establish the limiting principles of 4.675/2025 by injecting explicit consumer welfare and efficiency considerations. Bill 4.675/2025 should adapt to BCL, not the other way around. As Lazar Radic has written concerning similar amendments to the South Korean Fair Trade Act:
Introducing these exogenous principles for oversight of digital platforms into the existing antitrust statute might subvert the prevailing system’s logic by emphasizing bigness, structural presumptions, and sui generis rules in ways that could eventually spread to other sectors and undermine the integrity, consistency, and predictability of the law.
Organizational Challenges: The Digital Markets Superintendency
Under Article 5 of the BCL, CADE’s institutional design is tripartite:
- The Administrative Tribunal, composed of a president and six commissioners, serves as the agency’s adjudicatory body and issues final decisions;
- The General Superintendency (SG), headed by the general-superintendent, functions as CADE’s investigative arm and may take certain decisions of lower relevance (e.g., approving unproblematic mergers or dismissing early-stage investigations); and
- The Department of Economic Studies (DEE), led by the chief economist, provides economic analysis to both the Tribunal and the SG.
CADE’s president, the six commissioners, and the general superintendent are all appointed by the Brazilian president after being approved by the Senate in a public hearing. (The chief economist is jointly appointed in a decision by the general superintendent and the president.)
In essence, the agency’s investigative body (the General Superintendency) is separate from its adjudicatory body (the Administrative Tribunal), creating a system of checks and balances between the two bodies. While the Tribunal frequently follows the SG’s recommendations, it is not uncommon for it to overturn or modify them. This separation has long been regarded as a strength of the Brazilian system, preserving both procedural fairness and technical rigor.
Bill 4.675/2025 now proposes to create a new Digital Markets Superintendency as the investigative body charged with enforcing ex-ante rules for digital markets. The proposal to create an entirely new superintendency, rather than a specialized unit within the existing SG, surprised the Brazilian antitrust community. By establishing two parallel investigative bodies (one for ex-ante regulation of digital markets and the other for traditional ex-post enforcement), the bill risks introducing overlaps, inconsistent analyses, and even conflicting decisions.
Divergence between the SG and the Digital Markets Superintendency on issues like unilateral conduct or merger effects, for instance, could produce legal uncertainty and chill legitimate business conduct, ultimately harming the very innovation the law seeks to foster. In this sense, it is clear that Bill’s Article 14-B, § 2º and § 3º, which only states that collusive conduct and mergers will continue to be analyzed by SG, are insufficient to impede contradictory decisions, especially if the two superintendents have different views of antitrust enforcement against unilateral conduct.
International experience would appear to recommend a more integrated approach. The UK DMCC created a new Digital Markets Unit (DMU) within the Competition and Markets Authority (CMA) as a directorate alongside other enforcement functions, all reporting to the CMA’s CEO. Similarly, in the EU, the Directorate-General for Competition (DG COMP) handles DMA cases through a Deputy Director-General for Antitrust & Regulation of Digital Platforms. Placing a dedicated digital markets unit within CADE’s SG, rather than creating a separate superintendency, would be more consistent with Brazil’s institutional design.
Finally, the bill is silent on staffing and resourcing for this new body, in part due to the absence of a regulatory-impact assessment. But if economics teaches us anything, it is that there are always costs and tradeoffs. According to CADE’s 2024 Annual Report, the agency currently has 569 employees, of which roughly 55% work directly on CADE’s primary function of enforcing the BCL. Without clear authorization for new hires, CADE would likely be forced to reallocate personnel, potentially weakening its ability to handle cartels, mergers, monopolization cases, and other low-hanging fruit in industries with much more salient market power problems. This might include the fuel market, which CADE recently established as the priority industry for antitrust enforcement for the next two years.
In any event, if new positions must be created, which appears likely, a second legislative process will be required, thereby prolonging implementation.
Conclusion
Bill 4.675/2025 has only just begun its legislative journey, and much can still change before it becomes law. In the coming days, the president of the Brazilian House of Representatives will decide which committees will review the bill before it reaches the full House. Each committee will have the power to amend the text—potentially in significant ways. And this is just the first stage: if approved in the House, the bill will face a similar process in the Senate.
In other words, the text unveiled last Friday is far from final. It could emerge from Congress with major changes, minor adjustments, or virtually intact.
Politically, the outcome is uncertain. Although this bill enjoys government backing, it is not at the top of the administration’s priority list. President Lula’s government is currently consumed with such contentious issues as an amnesty bill for the Jan. 8, 2023 protesters, ongoing tax-reform debates, and budget constraints. Furthermore, with the 2026 elections on the horizon, legislative momentum is likely to slow, leaving the bill’s fate unclear.
Interestingly, early rumors suggested that the government would introduce two digital bills simultaneously: this antitrust proposal and a separate, politically sensitive content-moderation bill. The latter was ultimately shelved, likely due to the risk of defeat in Congress. In that sense, Bill 4.675/2025 will serve as a political test of the government’s ability to pass digital market regulation before it dares revisit the more controversial content-moderation agenda.
From a substantive standpoint, the “Brazilian DMA” could have been far worse had it more closely mirrored the original DMA’s prescriptive, one-size-fits-all model. By taking a more flexible, case-by-case approach—more similar to the UK’s DMCC, which is reported to be the bill’s primary international inspiration—Brazilian lawmakers are requiring CADE to conduct in-depth proceedings before designating companies as having “systematic relevance in digital markets” and imposing obligations on any company. This is an important safeguard that ensures regulatory obligations are economically and individually justified and better tailored.
Still, the bill’s reference to the ambiguous “competitive process” goal and the lack of a cognizable efficiency defense is likely to create friction—or worse, outright clashes with existing competition law.
Ultimately, however, the government has yet to make a convincing case that such a regime is even needed at the moment. The Ministry of Finance’s 2024 report relied on the familiar pro-intervention arguments: the bigness of tech companies, the perceived slowness of ex-post enforcement, “winner takes all” dynamics, and the inherent tendency of digital markets toward concentration. But even if these general hypotheses were correct in the context of Brazil, which is uncertain, it should be remembered that the nation’s existing legal framework requires regulatory impact assessments for agency rulemaking.
A cost-benefit analysis should likewise be conducted before adopting an ex-ante regime of this magnitude. Without that, there is no clear evidence that the bill’s benefits to consumers would outweigh the risks of overenforcement, reduced innovation, and higher compliance costs. In the absence of this assessment, there is a real possibility that the cure might be worse than the alleged disease.
Bill 4.675/2025 is set to trigger months and perhaps years of legislative debate. Regardless of one’s position, this is a defining moment for the future of digital-competition policy in Brazil. If lawmakers get it right, they could modernize the country’s antitrust toolkit while safeguarding innovation and consumer welfare. If they get it wrong, Brazil risks creating a costly and uncertain regulatory regime that fails to deliver on its promise. As Brazil steps into the global debate on digital regulation, the challenge will be to strike the right balance: protecting competition without punishing success.
