Lula’s campaign takes alleged Meta favouritism towards Bolsonaro to the UN and federal police

Lula’s coalition is challenging Meta over alleged unequal treatment in Brazil’s presidential runoff. It submitted a technical report to the UN Special Rapporteur on freedom of opinion and expression. The coalition argues that digital platforms must not distort the conditions under which voters form their views or compare candidates.

It also asked Prosecutor General Paulo Gonet and Brazil’s Federal Police to investigate whether Flávio Bolsonaro received economic advantages through Meta’s content-distribution systems. The filing requests statements from Bolsonaro, his campaign treasurer, and Meta executives in Brazil. It also asks officials to preserve Meta’s internal communications and seek records from the company’s US parent through judicial cooperation.

These steps do not establish that wrongdoing occurred. Meta attributes the reported incidents to technical errors, and Bolsonaro’s campaign had not responded on record. The investigation could determine whether the incidents were mistakes, systematic unequal treatment, or an electoral crime.

Meta is the company identified in the article as the owner of Facebook, Instagram, and WhatsApp. These are major digital communication platforms used for publishing posts, sharing messages, finding channels, and creating or viewing images. The company’s systems therefore influence how political material reaches users.

The reported allegations involve several different Meta products. Lula’s campaign said his Instagram profiles received reduced reach. It also said WhatsApp searches for the “Lula’s Army” channel suggested “Flávio’s Troop” instead. Meta’s artificial intelligence tool reportedly refused to generate images supporting Lula while generating images for his rival.

Meta attributes the incidents to technical errors. Even so, the dispute matters because platform behavior can affect the digital contest between candidates. The coalition asked investigators to examine whether these incidents created an economic advantage for Bolsonaro and whether Meta executives or company records can clarify what happened.

Reduced reach can make a candidate’s posts appear to fewer followers or users. That limits how widely messages, arguments, and campaign announcements circulate. A search system can also shape attention by placing one candidate’s channel before another when users look for political information.

The article gives specific examples. Lula’s campaign reported reduced reach for his Instagram profiles. On WhatsApp, searches for “Lula’s Army” allegedly suggested “Flávio’s Troop” instead. Meta’s AI tool reportedly refused to create images supporting Lula but generated images for Bolsonaro. Each incident could redirect attention or reduce a campaign’s ability to communicate.

The article does not establish that these outcomes were intentional. Meta says the incidents resulted from technical errors, while El País reported that Meta employees acknowledged restrictions on Lula’s account reach. Investigators would need to determine how often the effects occurred, why they happened, and whether they produced a meaningful imbalance.

Meta’s role in the article is broad because the company owns Facebook, Instagram, and WhatsApp. Together, these services cover public posts, visual content, private messaging, channel discovery, and artificial intelligence tools. The article does not provide user numbers, reach statistics, or market-share figures for Brazil, so its exact scale cannot be measured from the supplied text.

The reported incidents show several ways that role could matter politically. Lula’s Instagram reach was allegedly reduced. WhatsApp searches allegedly directed users toward Bolsonaro’s channel. Meta’s AI tool allegedly treated requests for candidate-supporting images differently. These functions can influence what campaign material people encounter or create.

The current dispute is therefore about platform power as well as individual technical failures. Meta says the incidents were technical errors, and the coalition is seeking evidence to test that explanation. Their electoral significance would depend on the incidents’ size, duration, causes, and effect on voters.

Campaign funding is not limited to money handed directly to a candidate. A valuable service, such as preferential distribution of political content, could potentially provide a campaign with an economic benefit. That is the theory behind Lula’s coalition asking whether Bolsonaro gained an advantage through Meta’s systems.

The filing asks the Prosecutor General and Federal Police to determine whether Bolsonaro received such an advantage. The reported examples include reduced reach for Lula’s Instagram accounts, a WhatsApp search suggestion favoring Bolsonaro’s channel, and different responses from Meta’s image-generation tool. If these effects were deliberate and valuable, the coalition says they could constitute an electoral crime.

The article presents this as an allegation, not a proven violation. Meta attributes the incidents to technical errors, and Bolsonaro’s campaign had not responded on record. Investigators would need to establish intent, unequal treatment, the benefit’s value, and whether Brazilian election law covers that form of support.

Lula’s coalition is seeking evidence that could distinguish accidental failures from unequal treatment. It points to a technical report and to internal messages, reported by El País, in which Meta employees acknowledged that Lula’s account reach had been restricted. Those messages could help establish what happened and when.

The filing asks Bolsonaro, his campaign treasurer, and Meta executives in Brazil to give statements. It also requests preservation of Meta’s internal communications. Finally, it seeks judicial cooperation from the United States to obtain records from Meta’s parent company. Such records may be held outside Brazil or involve systems managed beyond the local operation.

The requested evidence could clarify whether Bolsonaro’s campaign knew about or benefited from the alleged changes. It could also show whether the incidents affected many users, resulted from a shared technical problem, or reflected a decision. The article does not say investigators have obtained these records or proved wrongdoing.

In general, social-media systems use automated rules and learned models to rank content, predict what users seek, and decide which results appear first. They may consider signals such as relevance, prior interactions, safety rules, and technical settings. AI tools generate responses or images from user requests and the patterns learned from their training and controls.

The article shows why these mechanisms matter in elections. A distribution system may reduce the reach of one candidate’s posts. A search system may suggest a rival’s channel. An AI tool may refuse one candidate’s supportive image while creating one for another. Users can then receive different political information without seeing the underlying system’s decisions.

Meta says the reported incidents were technical errors. That explanation still matters because unintentional failures can affect equal campaign conditions. The coalition argues that election integrity includes voters forming views without undue manipulation or digital imbalance, so investigations must examine both impact and cause.