Meta Employees Allege Company Interfered in Brazil Election

Meta employees have accused the company of interfering in Brazil’s recent election by restricting President Luiz Inácio Lula da Silva’s social media accounts while favoring his challenger, a member of the Bolsonaro family. The allegations, made through a UK-based union, have been echoed by Brazilian prosecutors, the country’s top electoral court, and even Meta’s own Oversight Board, prompting legal action and public protests. Meta has denied any political bias, attributing the issues to technical failures and automated errors.

The denouncements came from employees represented by the Communication Workers Union (CWU) in the United Kingdom. In a statement, the union expressed concern over "obvious" and "biased" favoritism on Meta's platforms—Facebook, Instagram, and WhatsApp—during the runoff election on October 25. The employees alleged that "favours made to politicians such as the Bolsonaros have been programmatically made user-facing to such an obvious and biased extent."

Allegations of Suppression and Favoritism

The core of the complaint involves specific actions that allegedly hindered President Lula's campaign while promoting his opponent, Senator Flávio Bolsonaro. According to internal communications reportedly seen by employees, Meta’s platforms recommended Senator Bolsonaro's profile and his WhatsApp channel to users. Simultaneously, the Instagram accounts of President Lula were "hindered by restrictions."

Lula’s campaign filed a complaint with Brazil's Superior Electoral Court (TSE) on October 15, 2026, detailing these issues. The campaign claimed that some of the president’s Instagram posts appeared without a share button and that his stories blocked users from reacting or sharing. They also alleged "unjustified reductions in reach" for both paid campaign materials and organic posts.

Further accusations centered on Meta's artificial intelligence tools. Brazil's Federal Prosecution Office noted that Meta's Muse AI agent was apparently disabled from generating images in support of Lula, while it authorized the creation of images related to Flávio Bolsonaro. This prompted the TSE to order Meta to ensure its AI tool applies identical criteria to both candidates. In his ruling, Justice Marco Aurélio Bellizze stated that Meta could not reduce the distribution of political content or restrict searches in a way that created an imbalance between the campaigns.

Broader Scrutiny and Systemic Failures

The accusations from employees and the Lula campaign were amplified by criticism from oversight bodies and civil society groups. On October 20, 2026, Meta's own Oversight Board expressed concern over "recurring enforcement failures" in the company's election integrity efforts. The board specifically cited the "overenforcement of Meta's fraud policies against Brazil's President Luiz Inácio Lula da Silva's ad accounts," calling it a breakdown that should have been escalated to expert reviewers.

Concerns about Meta's content moderation systems were also raised by human rights organizations. Amnesty International reported that it had successfully submitted "mock" ads containing election disinformation to Facebook, including false claims about election rigging and incorrect voting dates. According to the organization, the ads were accepted even though the account used to place them had not completed Meta's identity confirmation process.

On October 23, 2026, a group of United Nations experts warned that "reports of restrictions on legitimate political expression on Meta platforms during Brazil's elections raise concerns for freedom of expression." A coalition of Brazilian civil society groups, including Article 19 Brazil and Sleeping Giants Brazil, stated they are documenting cases of takedowns, reach suppression, and algorithmic discrimination that appear to disproportionately affect progressive and democratic accounts.

Meta's Response and Official Actions

Meta has firmly denied all allegations of intentional interference. The company stated that "any suggestion that Meta favors one side is false." Instead, it attributed the problems to "technical limitations," "system failures," and "automated errors." According to Meta, the internal employee communications cited in reports related only to the severity of these technical issues, not to any deliberate political manipulation.

In a statement on October 20, 2026, Meta said it had "complied with an order of Brazil's electoral court for it to give equal treatment in its platforms" to both presidential candidates. The company also noted it had implemented a dedicated human review system for the campaigns' ad accounts. Meta defended its broader commitment to election integrity, stating it has quadrupled its global security and integrity team to over 40,000 people since 2016 and invested nearly US$5 billion in these areas in 2021 alone.

The official response from Brazilian authorities has been swift. On October 23, 2026, the Federal Prosecution Office demanded urgent answers from Meta about its content moderation policies, calling the company's previous responses "generic and insufficient" and giving it 48 hours to comply. The controversy also sparked public demonstrations, with a protest held outside Meta's London headquarters on October 24 and a larger one in front of its São Paulo office earlier that week. Lula's campaign has announced its intention to file a complaint with the Organization of American States (OAS), arguing that Meta manipulated social media during the election.

The central challenge in resolving these claims is the lack of transparency. According to João Victor Archegas, a coordinator at the Institute for Technology and Society, it is "almost impossible" to prove any interference by Meta because only the company has access to its internal data. This opacity limits the ability of external auditors, researchers, and government bodies to independently verify what happened.