TSE Ruling Forces Meta to Explain Its Algorithms to Brazil's Electoral Court

By Amara Osei

Brazil's Superior Electoral Court (TSE, the country's top electoral tribunal) has ordered Meta to impose no restrictions without public and equal criteria, require human review of advertising account blocks, and explain how its artificial intelligence operates. The ruling, issued by Justice Marco Aurélio Bellizze, came in an electoral judicial investigation lawsuit (AIJE) filed in October by the legal team of President Luiz Inácio Lula da Silva's campaign.

Angelo Ferraro, founding partner of Ferraro e Novaes Advogados and coordinator of the legal nucleus of Lula's 2026 campaign, laid out the stakes in an opinion column for O Globo. His core claim: electoral disputes now unfold on terrain that candidates do not control.

What did the ruling actually change?

Bellizze's decision, as Ferraro recounts it, required Meta to:

• Prevent restrictions applied without public and isonomic (equal-treatment) criteria
• Mandate human review of blocks on advertising accounts
• Correct flaws the company itself acknowledged
• Provide clarifications about Meta's artificial intelligence and investigate a network of 255 fake profiles

The campaign's lawsuit targeted three problems: distribution of content, blocked boosting (paid amplification of posts), and the activity of fake profiles.

Whose reach was cut?

Here the column is explicit. Ferraro states that Meta reduced the reach of profiles linked to Lula's campaign mobilization during the election period, and that the effects were asymmetric — damaging the candidacy while giving greater exposure to his opponent Flávio Bolsonaro.

That asymmetry claim carries the argument. Ferraro writes that the wave of TSE decisions established something relevant: Meta must explain its procedures, correct identified errors, apply equivalent criteria to all candidacies, and restore contracted reach that was never delivered.

A caveat the column does not address: these are the characterizations of one campaign's lawyer, not findings of an independent audit. The column cites no reach numbers, no deliverable-versus-contracted figures, and no Meta response.

Why does this end up in court?

Ferraro's answer is structural. Electoral justice works on short deadlines, while platforms concentrate data, tools and technical knowledge in their own hands. Without transparency and effective appeal mechanisms, he argues, the judicial route becomes the only way to clarify decisions that affect political communication — so litigation will tend to grow.

He frames the precedent in one sentence: electoral neutrality cannot be judged only by the content that stays online, but must also consider the rules determining who can speak, to how many people, and under what conditions.

The so-what is direct. If social networks have become a principal arena of political dispute, Ferraro writes, transparency about how they work is no longer a private matter for the companies — it is a question of public interest. Democracy, in his words, has entered the era of the algorithms; the question left open is how the TSE will police that power in the next elections.