Sanders and Democratic lawmakers ask Meta to explain alleged favouritism towards Bolsonaro

Bernie Sanders and 11 other Democratic lawmakers are challenging Meta over alleged unequal treatment of Brazil’s presidential campaigns. They want Zuckerberg to explain whether Instagram, Facebook and WhatsApp restricted or suppressed Lula-linked content while recommending or amplifying Bolsonaro-linked content. They call any candidate preference unacceptable, even if unintended.

The lawmakers also ask Meta to disclose whether it contacted Donald Trump’s administration about Brazil’s judiciary, platform regulation, the election or US policy toward Brazil. If such contacts occurred, they want records of meetings, calls and messages.

The request comes as Brazil’s Electoral Prosecutor’s Office says Meta failed to meet four equal-treatment requirements from the Superior Electoral Court. Meta denies interference and blames technical errors. The lawmakers’ demands could increase scrutiny of the platform’s election systems and its government contacts.

In this case, alleged favouritism means Meta’s platforms may not have treated both campaigns equally. Content linked to Lula could be restricted, suppressed or shown to fewer users, while Bolsonaro-linked material could be amplified or recommended more often. The lawmakers say this would be unacceptable even if it happened unintentionally.

The article gives specific examples. Lula’s campaign says his profiles had reduced reach, and Meta’s artificial intelligence refused to generate images supporting him while generating images for Bolsonaro. Bolsonaro’s campaign separately accused Meta of allowing negative ads against him and unlabeled AI-made Lula material.

These claims remain disputed. Meta denies interference and says technical errors caused the incidents, which it says were corrected. Brazil’s Electoral Prosecutor’s Office nevertheless says Meta failed four court-ordered equal-treatment requirements, keeping the alleged imbalance under official review.

Luiz Inácio Lula da Silva is identified in the article as Brazil’s president. Flávio Bolsonaro is identified as a senator. They are the two candidates facing each other in Brazil’s presidential runoff, the final contest scheduled for Sunday.

The article places their campaigns at the center of a dispute over Meta’s treatment of political content. Lula’s campaign says its profiles have suffered reduced reach and that Meta’s AI treated image requests differently. Bolsonaro’s campaign has also gone to court, raising complaints about negative ads and AI-generated material.

The contest is extremely close in the available polling. The article describes the candidates as holding a statistical tie and says they will take part in their final debate on Friday on TV Globo. It does not provide vote shares or explain the earlier election result that led to the runoff.

Brazil’s Superior Electoral Court is the institution whose election-related order governs Meta’s treatment of the two campaigns in this article. Its central requirement was equal treatment for Lula and Bolsonaro. That matters because platform decisions can affect how campaign messages circulate during the final stage of a presidential race.

The article says Meta had to comply with four points. They included providing statistics about its artificial-intelligence tool and handling complaints from both campaigns equally. The Electoral Prosecutor’s Office said Meta failed to comply with all four points and asked that the stipulated fine be applied.

The matter now rests with Judge Marco Aurélio Bellizze. He had warned that the daily penalty could rise to one million reais. The article does not list every one of the four requirements, so only the AI statistics and equal complaint handling can be identified specifically.

A statistical tie means the polling difference is small enough that the candidates’ positions overlap within the polls’ uncertainty. It does not prove they have exactly equal support. The article gives no vote percentages or margin of error, so the precise closeness cannot be measured from its figures.

Online reach could matter because recommendations and suppression determine which campaign messages users encounter. If one side’s posts appear less often while the other’s are amplified, voters may receive different information. Even a modest visibility difference could matter when the candidates are already closely matched.

The article does not establish that Meta changed the election outcome. It reports allegations from both campaigns, a court dispute and Meta’s denial. Still, the statistical tie helps explain why equal treatment is politically important before Sunday’s runoff and Friday’s final debate.

Systematic unequal visibility could affect what voters hear during the decisive days before the runoff. If one candidate’s messages were suppressed while the other’s were recommended, users might encounter an incomplete or skewed campaign debate. That could influence perceptions, even if the article does not prove an electoral effect.

The reported allegations illustrate the concern. Lula’s campaign says its profiles had reduced reach, while Bolsonaro’s campaign accuses Meta of allowing negative ads against the senator and unlabeled AI-made Lula material. The campaigns therefore claim different forms of unfair treatment, rather than one agreed explanation.

A perceived imbalance could also damage public trust in Meta, the election process and the eventual result. The article does not say that such a result has occurred. Meta denies interference and says technical errors were corrected, while Brazilian authorities are seeking compliance and possible penalties.

Content-moderation systems generally check posts against platform rules and may remove, label or limit material. Recommendation algorithms rank available content and decide what appears more prominently in feeds or suggestions. Generative artificial intelligence responds to prompts by creating material, or refusing when its safeguards block a request. These are general explanations; the article does not describe Meta’s internal rules.

The article connects those mechanisms to specific allegations. Lula’s campaign says its profiles had reduced reach, suggesting limited distribution. It also says Meta’s AI refused images supporting Lula while generating images for Bolsonaro. Bolsonaro’s campaign alleges that Meta allowed negative ads and AI-made Lula material without the required label.

Meta denies interference and attributes the incidents to technical errors it says it corrected. Brazil’s court nevertheless required equal treatment, AI-tool statistics and equal complaint handling. The dispute shows why transparency matters when automated systems affect political speech during a close election.