The Meta Advisory Council (Oversight Board) has expressed its concern about the "recurring" failures in the tech company's measures to protect the integrity of electoral processes in its services, warning that its current moderation policy is stifling political debate. This independent body of the company has specifically pointed out the recent technical limitations imposed on the accounts of the president of Brazil, Luiz Inácio Lula da Silva, just a few days before the second round of the presidential elections in the country.
In a public note, the co-chairs of this independent tribunal —responsible for reviewing decisions to remove or restore content on Facebook and Instagram, but not on WhatsApp due to its encryption system— have emphasized that Meta's errors are concentrated on two fronts: on one hand, an oversized application of the rules that ends up cutting off political conversation and, on the other, a lack of control that jeopardizes the cleanliness of electoral processes.
In relation to Brazil, the Advisory Council has warned that the excessive application of anti-fraud policies on Lula da Silva's advertising accounts constitutes "a failure that should have been referred to an internal team of expert reviewers."
Likewise, the body has criticized the absence of public indicators that allow for a clear measurement of the effectiveness of Facebook's matrix tools. Therefore, it urges Meta to improve its automated systems to integrate language, nuances, and political context, in order to avoid "irreversible damage" to citizen participation, something that, it emphasizes, has already been evident in the protests in Iran, Iraqi Kurdistan, or the United States.
Compliance with Brazilian Justice Orders
The positioning of this supervisory body comes in parallel to the recent judicial confrontation between the Brazilian Executive and the multinational. This Tuesday, Meta announced that it has completed the implementation of all provisions imposed by the Superior Electoral Court (TSE) of Brazil to ensure balanced treatment between Lula da Silva's campaign and that of his opponent at the polls on October 25, the far-right senator —and son of former president Jair Bolsonaro— Flávio Bolsonaro.
The Brazilian electoral Justice had set a deadline for the company, under the threat of a daily fine of 500,000 reais (about 87,000 euros), after a complaint from the government coalition. The candidacy of the Workers' Party (PT) accused the artificial intelligence tool 'Meta AI' of refusing to generate campaign material for the current president, while it did accept similar requests in favor of the conservative candidate. Furthermore, Lula's team reported that their ads on Instagram were blocked for several days, a problem that the company initially attributed to a "global failure."
To comply with the ruling of Judge Marco Aurélio Bellizze, Meta has launched a new system specifically designed for these elections, claiming the technical complexity of accurately identifying all paid political advertising. The company has insisted that this error affected "only advertising and not organic posts" and has reiterated its commitment to comply with the TSE's decisions in order not to interfere with the normal conduct of the elections.