Meta Ignores Brazilian Court Order to End Bolsonaro Favoritism

Meta's AI search and content suggestion tools may manipulate election outcomes, and this court case could force platform transparency.

Two civil society groups, Sleeping Giants Brasil and CTRL+Z, have presented new evidence to Brazil's Superior Electoral Court alleging that Meta's platforms, specifically Instagram and WhatsApp, exhibit unequal treatment favoring Senator Flávio Bolsonaro over President Luiz Inácio Lula da Silva. The organizations claim search results for GIFs and hashtags, as well as Meta AI responses, disproportionately promote Bolsonaro or link Lula to negative content. This evidence was submitted despite a prior court order demanding Meta ensure equal campaign conditions. The groups are requesting the court to enforce a daily fine of R$500,000 (approximately $99,000) and compel Meta to explain its AI-driven ranking and response mechanisms. The court has already warned Meta that the fine could double if non-compliance continues, demanding detailed explanations and action on the alleged algorithmic bias.

The core issue is not simply the presence of political content, but how automated systems on Meta's platforms organize and present information during a critical election period. The organizations are pushing for Meta to disclose the specific criteria behind its algorithms for ranking images, suggesting hashtags, and generating AI responses, aiming to uncover potential systematic biases.

This legal challenge highlights a growing global concern about the influence of AI and platform algorithms on democratic processes. The Brazilian court's insistence on transparency and equal treatment could set a precedent for how social media companies are held accountable for the societal impact of their automated content curation systems.