On Friday, Meta pushed back against allegations that its platform had shown political favouritism during Brazil's presidential race, with the response arriving within hours of a ruling by the country's electoral court directing the company to provide equal treatment to both Luiz Inácio Lula da Silva and Flávio Bolsonaro as they prepare for their Oct. 25 runoff.
Legal action against the platform operator behind Instagram, Facebook, and WhatsApp stemmed from allegations put forward by Lula's campaign team, which contended that the company's algorithms and artificial intelligence tools were configured in ways that gave Bolsonaro an advantage.
In a statement, Meta attributed the string of platform disruptions to a combination of unconnected worldwide technical faults, anti-fraud protocols, and obligations tied to regulatory compliance, rather than to any deliberate effort to suppress content. The organisation maintained that its platforms deliver content without regard to political viewpoint, adding that enforcement guidelines apply uniformly across all users.
Two separate platform-wide outages were flagged by the company as having taken place earlier in October. The first left view tallies on Instagram Reels frozen across more than 100 countries for a period while content itself continued reaching audiences normally, and the second knocked out ad creation from existing Instagram posts within one of its tools for a single day.
In responding to concerns raised by the Lula campaign over Meta AI, the platform noted that a gap in its filtering system — triggered when generic first names were paired with electoral identification numbers — had temporarily blocked certain users from producing images of the president, while that same gap left image generation for Flávio Bolsonaro accessible for an additional day.
The company's response left unaddressed a question about why Bolsonaro's Instagram and WhatsApp accounts were served algorithmic recommendations — a practice that Brazil's electoral court had ruled to be unlawful — while a spokesperson argued that voters deserved factual information rather than a narrative built by linking unrelated technical matters.
In a statement, Meta outlined its intention to uphold platform guidelines, operate within Brazil's electoral legal framework, address any errors as they arise, and maintain consistent treatment across all users, adding that candidates, parties, and political viewpoints in the country are subject to the same standards applied to its platforms during elections elsewhere around the world. Content moderation, the organisation stated, operates without favouring or penalising any particular viewpoint.
The statement went on to say that no political bias was present in any of the incidents the Lula campaign had pointed to. Internal communications obtained by a prominent Brazilian newspaper pointed to post-promotion restrictions on the incumbent campaign's accounts, with platform staff tracing those limitations to system malfunctions and automated processes rather than deliberate enforcement decisions.
Three measures addressing digital safety — protective rules for women and minors, decrees placing accountability on platforms, and a joint arrangement with the Supreme Court to pursue companies over prohibited content and false information — took shape under Lula's administration following his 2023 inauguration. Under Brazilian law, Google, Meta, and TikTok have faced obligations since last year to actively oversee their platforms and take down content involving hate speech, racist material, or calls for violence. That stance has introduced friction into the administration's ties with the White House, and drawn pushback from those who contend it threatens freedom of expression.
In a statement, Meta indicated its intention to keep mounting a defence against the lawsuit.
The lawsuit filed in Brazil draws on a pattern also visible in Romania and Germany, where judicial bodies have stepped in to address disparities in how platforms operate. Ahead of the 2019 European Parliament elections, a ruling handed down by Germany's Constitutional Court required Facebook to reinstate a political party's removed page, with the judges holding that large technology platforms are bound by constitutional guarantees of equal treatment in the political sphere.
Investigators determined that Romania's 2024 presidential vote had been subjected to undisclosed algorithmic promotion through TikTok and coordinated bot activity that directed attention toward candidate Călin Georgescu, a finding that led authorities to void the result and pass emergency measures mandating identification codes on political content distributed across social media platforms.
Researchers at the Federal University of Rio de Janeiro published findings this week documenting that the platform hosted no fewer than 1,613 political advertisements deemed non-compliant with electoral rules — accumulating 171 million views and approximately 2 million reais (around $400,000) in revenue across the July-to-September window — with the bulk of that funding traced back to sources in the United States. Under Brazilian electoral law, purchasing such advertisements is restricted exclusively to registered political parties, candidates, or formally recognised coalitions.
Santini, a communications scholar who contributed to the study, argued that Meta ought to be barred from accepting political campaign advertising — whether the buyer is based domestically or overseas — and pointed out that the platform has a record of running into electoral law disputes across multiple countries. Santini went on to describe the pattern as one rooted in the company's broader operational conduct.
