The impact that the elections could have on the Brazilian electoral scenario is still quite uncertain. There are several gaps in the way of implementation and little transparency about how the company already applied its current rules.
Although he has promised to go back on the policy of reducing the reach of political posts (called “civic content” in the company), it is difficult to predict whether this will result, for example, in an even greater radicalization of the digital debate or whether any political field can end up benefiting more from new changes to the algorithm.
Meta owns Facebook, Instagram and WhatsApp. Last Tuesday (7), Zuckerberg released a video announcing changes to the company’s content moderation policies, in what was seen as a nod to US President-elect Donald Trump. The decision includes the end of the fact-checking program — now it will be users who will include corrections in posts that may contain false information, similar to what happens on X (formerly Twitter).
The impact of implementing community notes will also depend greatly on how it is put into practice — if this happens in Brazil.
Among the open points about the changes is how this new system would be created, who would make it up and how ideological biases would be mitigated.
Furthermore, another aspect is the type of content for which this system would be valid. Today, for example, posts from politicians and candidates are not within the scope that can be analyzed by checking agencies, for example.
Another relevant aspect is that there is also no transparency about which public figures are in .
A scenario that, for now, is emerging with greater concreteness is that, faced with the removal of part of the rules on gender, vulnerable groups face a more violent digital and campaign environment.
One of the updates now allows, for example, .
The TSE (Superior Electoral Court) resolution on electoral propaganda, on the other hand, brings strings attached, in a more specific way than in the rest of the legislation, providing, for example, the obligation for companies to adopt effective measures to mitigate certain risks, including gender-based political violence. It also establishes the need for content policies that seek to reduce the circulation of misinformation.
Another measure announced by Meta that could have an impact is the removal of automated detection of “low-severity violations”, which would then depend on user reports for analysis. It is unclear, however, what would fall into this category.
The company’s decision goes against what experts say, who point out not the need to reduce automation, given the volume of posts every second, but to increase human review teams, with knowledge of the language and context from each country – the company does not reveal how many such teams it currently has.
Viktor Chagas, who is a communications professor at Universidade Federal Fluminense, states that, due to the lack of transparency in networks, it is a great challenge to try to measure how content removal and reach reduction policies are carried out.
“It has always been a big black box, we have always had extreme difficulty measuring how this type of moderation policy is applied,” he says. “The very definition of what political content is, among researchers themselves, would certainly not be such a simple consensus.”
Mariana Carvalho, professor at FGV Comunicação Rio, assesses that it will still be necessary to understand how the new rules will impact moderation, but predicts that combating disinformation will be more difficult. She highlights, on the other hand, the difficulty in carrying out studies on the impact of changes, such as .
Regarding the restriction on content about politics, she says that it is difficult to measure how much the limitation made sense during the electoral period, when there is great interest in the subject.
Last year, this policy was listed, with emphasis on the aspect of transparency, among the reasons for opening an investigation. In 2022, when it was already in force, Brazil experienced a very radicalized campaign, which would culminate in the attacks of January 8th.
One of the effects already announced in the USA, and which should be replicated elsewhere in the world, is that .
In 2024, this possibility was already available in the settings. But during the election period due to a rule, stipulating that they could not be recommended.
Carlos Affonso Souza, director of ITS (Institute of Technology and Society) and lawyer, considers that it is still too early to understand how much what was announced will affect the electoral debate. On the other hand, he emphasizes that, with the TSE’s resolutions, there is already a better path outlined for dealing with the company.
Among the uncertainties, he highlights that it is not known whether low-severity violations — and which would depend on user reports for moderation — would include content about electoral integrity and produced by artificial intelligence. And he bets that, by 2026, the TSE should create rules on the use of community notes for moderation, just as it dealt with artificial intelligence last year.
A point of attention is what will be the stance on the issue of minister Kassio Nunes Marques, appointed by Jair Bolsonaro (PL) to the STF (Supreme Federal Court) and who will be the next rapporteur of the process of updating electoral resolutions and president of the TSE during the 2026 election.
Also, the eventual completion of the project could place more barriers to what Zuckerberg promises to change.
Asked about the application of the measures in Brazil, whether the restrictions on posts by politicians for checking would be repeated in the community notes model and about the existence of reports on how it has been applying its policy on political content, Meta restricted itself to sending the last week.
