We need to talk about Meta: Political interference and interests in the Brazilian elections

On the eve of Donald Trump’s return to the presidency of the United States, in January 2025, Meta’s founder and CEO, Mark Zuckerberg, announced that changes would be made to his company’s moderation policies in the name of “freedom of expression”, which, he said, was under threat by global pressures for what he saw as censorship. The announcement signaled a rapprochement between Meta and the far right around a common goal: the deregulation of Big Tech. Since then, growing entanglement between major technology companies and the U.S. President Donald Trump administration has shown how the private interests of a small group can override the public interest.

In recent months, allegations have been mounting that Big Tech companies, especially Meta, have been interfering in the Brazilian electoral process this year, aiming to undermine the reelection of President Lula (PT). This kind of accusation is not new. Over the past decade, a vast body of literature has exposed the mediating role played by digital platforms in foreign interference in elections around the world. These reports have also highlight the challenges of holding companies accountable for allowing these interference campaigns to be amplified.

At our research group, NetLab UFRJ, we have spent years investigating how major digital platforms shape public debate and influence people’s perceptions of reality, often outside the regulatory and normative frameworks in force in Brazil. Meta in particular has become fertile ground for disinformation and boosting fraudulent ads, which already account for 10% of its annual revenue.

Despite lawsuits and administrative actions brought by Brazilian authorities against Meta, complaints and requests from civil society for investigations, as well as reports published by the press, the company has shown neither the willingness nor the diligence needed to effectively fight this kind of content.

Given this troubling track record, NetLab UFRJ created the “Big Tech Observatory” during the 2026 election campaign period in order to monitor how Meta and other digital platforms performed in two key areas: algorithmic recommendation and political advertising. The initiative set out to examine whether these companies would comply with the current Brazilian legislation established by the Electoral Supreme Court (TSE). The results of the analyses, however, told a predictable story.

We found that, between July and September, Meta received approximately R$ 2 million (US$ 400,000) for running 1,613 ads placed from outside Brazil that aimed to either favor or harm candidates. All were paid for in foreign currencies and boosted by pages with declared locations in other countries. The amount invested in this advertising is equivalent to that spent on prominent campaigns, including the race for Presidency and other executive offices.

Brazilian electoral law expressly prohibits candidates from receiving cash or donations in-kind from abroad. This ban, however, did not prevent the ads from being displayed to Brazilian users of Meta’s platforms. According to the available information, not only the ads were allowed to circulate but they may have reached up to 171 million views in the country.

But the irregularities in advertising do not stop there. Brazilian electoral rules prohibit electoral advertising from 48 hours before the polls to 24 hours afterwards. However, our analyses of both rounds of the 2026 election identified hundreds of ads in favor or against specific candidates that circulated on Meta’s platforms with no apparent restrictions. The company even allowed fake profiles posing as official channels of the Electoral Supreme Court to boost disinformation about a supposed “digital vote” which stated that people would be allowed to cast their vote from home on the day of the first round — something which does not exist in the Brazilian electoral system.

In addition to these irregularities, Lula’s political campaign page was temporarily disabled by Meta without explanation, preventing it from running electoral ads. On September 23, his Worker´s Party (PT) reported they had been blocked from advertising on Meta’s platforms. In a statement released on October 16, the company rejected the accusations of election interference and attributed the temporary deactivation to an “incorrect anti-fraud review” carried out under an agreement with the federal government to fight fraud involving public programs.

The effectiveness of the company’s anti-fraud measures, however, is questionable: an analysis carried out by NetLab UFRJ identified hundreds of fraudulent ads involving federal public policies circulating before and after September 23, indicating that those same measures were not able to stop this kind of content from running during the same period.

Finally, we found signs of discriminatory practices in ad pricing. Based on data released by Meta, we estimate that for the first round Lula’s campaign may have paid between 33% and 43% more than his far right adversary Flávio Bolsonaro campaign to reach the same number of users on Meta’s platforms. Despite this substantial difference, limitations in the company’s transparency policies make it impossible to identify precisely which parameters and criteria are used to price ads and, consequently, to draw definitive conclusions about the mechanisms behind this disparity. Even so, studies conducted in other contexts, such as the United States, India and the Netherlands, also point to evidence of discriminatory pricing practices in the platform’s advertising.

Other complaints made during the elections add to this overall scenario of negligence and show that, despite being worth billions and their strategic position at the forefront of technological innovation, major digital platforms invest little in protecting their users. For example, this 2026 election in Brazil saw several controversies involving WhatsApp’s image generation service, powered by the Meta AI artificial intelligence. Users reported being able to generate content favorable to Flávio Bolsonaro’s campaign without difficulty, whereas they were not getting the same results for Lula. Meanwhile, the documentary channel Revelando had videos criticizing Flávio Bolsonaro removed from its Instagram profiles shortly after they were posted, without clear justification from the company.

These episodes cast doubt on the consistency of Meta’s moderation and content control systems, exposing failures that call for greater investment and improvement. The situation is made worse by the absence of transparent criteria or guarantees of equal treatment in how ads are handled, how content is recommended, how advertising is priced, and how profiles and posts are moderated.

Big Tech companies have developed predatory business models, and they are currently joining forces with political forces committed to deregulation in Brazil. This is part of a strategy to protect their businesses and expand their room for maneuver.

Whether or not one side of the electoral race is being deliberately favored, it is indisputable that the major digital platforms continuously exercise the power to determine what gains visibility and what remains on the margins, with no transparency over their criteria and no mechanism for accountability. What is clear is that, for these platforms, defending their economic interests and maximizing their profits prevails above user rights, public interest, and democracy itself.

Os autores não prestam consultoria, trabalham, possuem ações ou recebem financiamento de qualquer empresa ou organização que se beneficiaria deste artigo e não revelaram qualquer vínculo relevante além de seus cargos acadêmicos.