TSE Fines Meta R$7.5 Million and Exposes the Paradox of Electoral Transparency

Brazil’s electoral judiciary demands verifiable evidence from Meta about how its systems operate. But how far does the autonomy of those overseeing technologies administered by the TSE itself extend? The fine imposed on Facebook reveals an institutional question that reaches beyond digital platforms to the limits of oversight of public power.

Fernando Fagundes Ferreira ·

There is a fundamental difference between stating that a system operates impartially and providing independent means of verifying that claim. The principle took on particular significance on November 10, 2026, when the Superior Electoral Court (TSE), by unanimous decision, imposed a fine of R$7.5 million on Facebook Serviços Online do Brasil, a Meta company, for failing to comply with court orders concerning the distribution of political content during the presidential runoff. The decision did not establish that the company had deliberately favored any candidate, but it found that incomplete or late compliance with the orders, accompanied by insufficient information, hindered the oversight sought by the court.

The controversy contains an institutional irony that goes beyond the platform’s conduct. The Electoral Court, responsible for administering Brazil’s electronic voting system, demands that a private corporation provide effective means of verifying its technological mechanisms. The same principle, however, must guide public evaluation of the systems used by the State itself. Brazil has mechanisms for external oversight of electronic voting machines, with the participation of political parties, oversight bodies, and other authorized entities. Their existence makes it impossible to say that electoral verification takes place exclusively within the TSE, but it does not remove a relevant question: how much technical autonomy do overseers have to carry out, document, and reproduce their own checks, particularly given that some procedures take place in environments and under conditions established by the authority responsible for the system?

The question presupposes neither electoral fraud nor technical equivalence between an electronic voting machine and commercial content-distribution algorithms. These are technologies with different purposes, architectures, and risks, subject to different oversight regimes. The connection is institutional and forms the core of this analysis: a public authority that demands verifiable evidence from third parties must also demonstrate the quality, scope, and limits of the mechanisms through which its own actions can be examined. This requirement does not weaken democratic legitimacy. On the contrary, it shifts the basis of that legitimacy from the authority of those who administer the systems to the strength of the evidence that allows them to be assessed.

The fine and the power to control visibility

The complaint that led to the penalty was filed against Facebook by the Brasil Pronto Pra Mais coalition, which was associated with Luiz Inácio Lula da Silva’s candidacy. The campaign alleged differences in the treatment of electoral content, including restrictions on the reach of posts, problems involving political advertisements, and possible irregularities associated with automated systems. Meta disputed the allegations and provided information about the measures it had taken, maintaining that its policies were applied impartially. In the course of the proceedings, however, the reporting justice, Justice Marco Aurélio Bellizze, found that certain obligations had been met after the deadlines, incompletely, or with explanations insufficient to allow adequate verification of the required measures.

According to the ruling, the company provided some data in fragmented form, substituted its own timeframes for those set by the court, and relied on aggregated information or generic statements that made oversight more difficult. The penalty consists of three days of fines of R$500,000, from October 17 to 19, and six days of fines of R$1 million, from October 20 to 25, totaling R$7.5 million. The basis for the ruling was the failure to comply with the court orders, considered as a whole—not proof of intentional electoral interference. Bellizze emphasized that demonstrating deliberate conduct to unbalance the election would require robust evidence, to be produced through the appropriate legal process.

The distinction deserves attention because the power of digital platforms does not depend solely on the explicit removal of posts. Content recommendation and distribution systems can influence how often a message appears, which audiences it reaches, the relative placement of posts, and the conditions under which sponsored content circulates. A post can remain available and still experience a significant decline in visibility, without that necessarily constituting censorship or discriminatory treatment. Differences in reach may result from user behavior, advertising strategies, content characteristics, legitimate moderation mechanisms, or technical failures. The multiplicity of these explanations makes it essential to have information that allows ordinary variations to be distinguished from possible improper practices.

This ability to manage visibility represents an important transformation in the relationship between communication and political power. For decades, debate over the influence of the media focused on the ownership of newspapers and broadcasters, the selection of news, and the shaping of public agendas. Platforms have added an automated infrastructure for mediation, capable of organizing the circulation of information among segmented audiences according to criteria that combine corporate decisions, statistical models, commercial interests, and behavioral patterns. The existence of these mechanisms does not prove political manipulation, but it does establish a concentration of informational power whose exercise can have significant effects on democratic debate. Public oversight must therefore consider not only what platforms say they do, but what their systems actually do.

A company may assure the public that it treats all candidates equally, report that it has corrected failures, or present reports on its moderation policies. Those statements, however, do not necessarily replace the records and procedures required to verify compliance with specific obligations. The problem becomes more acute during election periods, when automated decisions can have effects over relatively short intervals, while judicial investigations depend on formal stages, submissions by the parties, and the production of evidence. The fine imposed after the end of the runoff demonstrates the possibility of holding a company accountable for failing to comply with court orders, but it does not, by itself, allow a reconstruction of all the informational conditions that would have existed had the orders been fully complied with and met on time.

Oversight autonomy and the limits of institutional power

The question of verifiability also deserves examination in the Brazilian electronic voting system itself, while respecting the technical differences between it and digital platforms. Voting machines do not select messages for voters, organize personalized recommendations, or use commercial engagement mechanisms like those of social networks. Their purpose is to record and count votes through a specific technological architecture, subject to preparation, authentication, oversight, and auditing procedures. The TSE regulates source-code inspections, checks of the integrity and authenticity of programs, tests conducted with randomly selected voting machines, and mechanisms for checking voting-machine result records, allowing the participation of legally authorized external entities.

The participation of external institutions is an important element of this structure. In August 2026, for example, technicians from the Office of the Comptroller General of the Union (CGU) took part in inspecting the source code of systems intended for the general elections, including programs related to voting, equipment preparation, file transmission, and the tabulation of results. Electoral regulations also permit oversight entities to develop their own verification tools and provide for independent vote-checking procedures in specific audits. These mechanisms demonstrate that oversight is not limited to Electoral Court employees and includes components that do not rely exclusively on tools supplied by the court.

The more relevant debate, however, concerns the extent of the autonomy afforded to oversight agents. Certain code inspections take place in a controlled environment and follow regulatory conditions for using analysis tools, recording information, and reproducing results. Verification programs developed by oversight entities are also subject to technical specifications and approval procedures. These requirements may address legitimate needs for system security and integrity, but their existence does not obviate the need to examine their proportionality and their effect on the capacity for independent investigation.

There is an important conceptual difference between institutional independence and methodological autonomy. An organization may not be part of the TSE’s administrative structure and still depend on environments, access conditions, or procedures defined by the court to carry out certain checks. That does not automatically invalidate its conclusions or reduce external oversight to a formality. It means that the quality of an audit must also be examined in terms of freedom to choose methods, the ability to document findings, the capacity to reproduce tests, and the conditions provided for specialists to subject results to technical scrutiny. The more important an infrastructure is to the public, the more important it becomes to clarify the scope and limitations of its verification mechanisms.

It is precisely at this point that the fine against Meta raises a question that goes beyond the legal dispute. By finding certain information provided by the company insufficient, the TSE reaffirms that oversight must create concrete conditions for verification rather than depend exclusively on statements by the organization being overseen. The comparison does not justify transferring the same technical procedures to the electronic voting system, but it does allow us to ask whether the conditions established by the court itself give external entities sufficient autonomy to identify relevant failures, document them adequately, and produce reproducible results. The legitimacy of this question derives from the very principle of transparency invoked by the Electoral Court.

This discussion does not require choosing between absolute trust and generalized suspicion. The security of a technological system can be supported by consistent evidence accumulated through different verification procedures, without precluding improvements to its oversight mechanisms. Likewise, the existence of access limitations does not justify concluding that the system is fraudulent or that electoral results have been manipulated. Democratic debate concerns the ability of institutions to explain their technical choices, justify restrictions, disclose results, and respond to substantiated questions without conflating them with unsupported accusations.

The Meta ruling itself offers an example of the tension between transparency and the protection of sensitive information. Although it required data intended to oversee compliance with electoral obligations, the TSE preserved the confidentiality of certain documents related to Meta AI’s filters and parameters for detecting synthetic content, recognizing their commercial sensitivity. The decision demonstrates that the requirement of verifiability does not necessarily entail unrestricted disclosure of code, models, security parameters, or protected information. The challenge is to establish procedures that enable independent assessments without unnecessarily compromising technological integrity, privacy, or other legitimately protected interests.

The same logic applies to public infrastructure. It is possible to preserve sensitive components of an electoral system while ensuring robust external audit mechanisms, technical documentation, and the disclosure of evidence. Solutions depend on the characteristics of each technology and the risks associated with its operation, but access restrictions must be justified and consistent with oversight objectives. Public trust should depend neither on unquestioning acceptance of a private company’s statements nor on a presumption that a state institution is infallible. It must rest on procedures capable of demonstrating how systems work, and the limits of the conclusions reached, according to known technical criteria.

Electoral sovereignty and informational power

The controversy involving Meta reveals a growing challenge for contemporary democracies: a significant part of public opinion is formed in environments administered by transnational private companies, subject to commercial interests, their own moderation policies, and different legal systems. These corporations do not determine voters’ behavior on their own, but they influence the conditions under which information, advertising, and political participation circulate. In this context, protecting electoral sovereignty depends on the ability of States to require compliance with legitimate obligations without turning platform regulation into an instrument for arbitrarily controlling public debate.

The risk exists in both directions. Inadequate oversight may leave corporate decisions that matter to the circulation of political information beyond public scrutiny. On the other hand, excessive concentration of regulatory powers in state authorities may produce disproportionate restrictions on freedom of expression, selective enforcement of rules, or undue interference in political communication. A democratic response requires accountability mechanisms that simultaneously safeguard corporate transparency, users’ rights, freedom of the press, due process, and the possibility of oversight of public decisions themselves.

The lawfare dimension lies in the need to examine the possible instrumentalization of legal and informational mechanisms in political disputes. The concept cannot be applied automatically to every investigation, sanction, or judicial decision with electoral consequences. Its analytical use requires evidence of selectivity, improper purpose, procedural manipulation, or the strategic use of law to achieve objectives inconsistent with its impartial application. The fine imposed on Meta, by itself, does not demonstrate that such practices occurred, but it offers a relevant case study of the interaction among technology companies, judicial institutions, political campaigns, and disputes over the legitimacy of information.

Journalistic investigation must hold all these actors to the same evidentiary standard. Allegations of platform interference must be substantiated, just as any claims of irregularities in voting systems must be supported by verifiable technical evidence. Judicial decisions, in turn, must be reasoned and subject to legal mechanisms of review. Criticism of the relationship between technology and power loses its rigor when it replaces documentary analysis with suspicion or turns possibilities into established facts. A commitment to transparency requires recognizing both what has been proven and the limits of what can still be investigated.

Justice Marco Aurélio Bellizze linked the influence of digital platforms to popular sovereignty, political pluralism, and voters’ autonomy, recognizing that the protection of political rights extends beyond the moment when a citizen casts a ballot. It also includes the informational conditions in which citizens form their convictions and take part in public debate. This concern is relevant, but popular sovereignty is not the exclusive prerogative of the authorities responsible for administering elections. The power to oversee platforms is exercised on behalf of society and within constitutional limits, through reasoned and proportionate decisions that are subject to review.

The same responsibility extends to the administration of technological systems used by the State. Publicly presenting the evidence that supports the security of Brazil’s electoral process, enabling meaningful participation by external entities, and continually improving audit mechanisms can help strengthen confidence in the process. These measures help distinguish legitimate technical questions from disinformation campaigns that make unsubstantiated accusations. An institutional culture of verification is not a threat to democratic legitimacy; on the contrary, it offers better conditions for protecting legitimacy from narratives built on claims that do not withstand factual scrutiny.

The November 10 ruling does not settle the dispute over possible electoral favoritism by Meta’s platforms. The decision found that court orders had not been complied with, but it did not establish that the company had deliberately manipulated the 2026 presidential election. Its importance, however, extends beyond the financial penalty because it makes explicit a requirement that should guide the relationship between technology, institutional power, and political rights. Meta has a duty to comply with legitimate orders and provide the elements needed for oversight. The Electoral Court, in turn, must continue demonstrating the robustness of its own procedures and ensuring adequate conditions for qualified external agents to examine them critically.

The existence of external oversight does not necessarily mean unrestricted methodological autonomy, just as the presence of audit mechanisms does not remove the need to assess and improve them. Recognizing these distinctions allows technological governance to be discussed without delegitimizing institutions or fueling unsupported suspicions. Democracy is strengthened not by the absence of questions, but by the ability to answer them with evidence, transparent procedures, and public accountability. The right to oversee must not become a prerogative to escape oversight, and institutional authority does not replace the need to demonstrate what it asserts.

Demanding trust is not enough. It is necessary to provide the means to verify it.

---

Sources and documentation

Tribunal Superior Eleitoral. TSE fines Facebook R$7.5 million for failing to comply with court orders. November 10, 2026. https://www.tse.jus.br/comunicacao/noticias/2026/Novembro/tse-multa-facebook-em-r-7-5-milhoes-por-descumprir-determinacoes-judiciais

Ministério Público Federal. 2026 Elections: Meta fined R$7.5 million for failing to comply with court orders. November 11, 2026. https://www.mpf.mp.br/o-mpf/unidades/procuradoria-geral-da-republica-pgr/noticias/eleicoes-2026-meta-e-multada-em-r-7-5-milhoes-por-descumprir-determinacoes-judiciais

Tribunal Superior Eleitoral. Resolution No. 23.673, of December 14, 2021. Oversight and auditing of the electronic voting system, as amended. https://www.tse.jus.br/legislacao/compilada/res/2021/resolucao-no-23-673-14-de-dezembro-de-2021

Tribunal Superior Eleitoral. CGU completes inspection of the source code of electoral systems for the 2026 Elections. August 21, 2026. https://www.tse.jus.br/comunicacao/noticias/2026/Agosto/cgu-conclui-inspecao-dos-codigos-fonte-dos-sistemas-eleitorais-para-as-eleicoes-2026

Tribunal Superior Eleitoral. 2026 Elections: source code of electoral systems remains open for inspection. April 14, 2026. https://www.tse.jus.br/comunicacao/noticias/2026/Abril/eleicoes-2026-codigos-fonte-dos-sistemas-eleitorais-continuam-abertos-para-inspecao

Tribunal Superior Eleitoral. Resolution No. 23.769/2026. Update to the rules on electoral auditing and oversight.

Judicial proceeding: Representation No. 0602746-69.2026.6.00.0000, Superior Electoral Court.