The Electoral Leviathan

The election campaign has begun, and along with it, a more equipped TSE (Superior Electoral Court) willing to decide what can or cannot circulate: a new resolution gives the Court the power to label, suspend, and hold accountable any content it classifies as “manipulated” by artificial intelligence — with immediate removal and the burden of proof reversed against whoever published it. The ruler that will measure truth from lies, in yet another election, rests in the pen of the robed justices occupying the seats of our higher courts.

Artificial intelligence is still a novelty. While many see it as positive and others as negative, everyone, without exception, must deal with the abrupt transformations it has caused across countless fields — and politics is no different. It’s natural to feel wary. Still, fear of the new cannot turn into a demand for more state control over our lives. Every new technology that frightens people generates calls for more control. Every round of control, if poorly designed, sows the seeds of the next abuse that will justify the following round.

The renowned economist Milton Friedman described this mechanism in many of his works: every state intervention creates distortions, and those distortions become the justification for the next round of intervention — a spiral that never solves the original problem, only pushes it forward with more power concentrated in the State. Of course artificial intelligence carries risks. Just as accessing the internet itself carries risks. Or just as driving a car carries risks. What Friedman pointed out is that, in most cases, the interventionist solution doesn’t ease the problem — it just creates other problems, of other kinds.

The risk of handing the power to decide what voters can or cannot read over to the political class is greater than most people can imagine. The situation becomes even more serious when this happens during an election period. The TSE’s new resolution is just one more tentacle of the Leviathan that controlled media outlets’ freedom of expression in the last presidential election. The problem isn’t just that a ruler exists. It’s which side it tends to lean toward — and the recent timeline shows a pattern that has repeated for four years, between the TSE and the STF (Supreme Federal Court).

In 2022, it fell to the TSE to repeatedly decide what Brazilians could say about Lula: Justice Tarso Sanseverino suspended campaign ads that called him a “thief,” even as a former STF justice publicly acknowledged that the Court had never cleared him. Weeks later, the Court ordered the removal of a Brasil Paralelo video under the label of “informational disorder.” During the same period, according to a survey by Gazeta do Povo, dozens of other requests from the PT (Workers’ Party) were granted by the Court.

Almost four years later, the pattern has returned — and gained a second front. In April 2026, after candidate Romeu Zema began satirizing STF justices in the series “Os Intocáveis” (“The Untouchables”), Justice Gilmar Mendes requested that he be included in the same fake-news inquiry that has already lasted seven years. In May, the Attorney General’s Office charged Zema with slander against Gilmar himself. In July, it was the TSE’s turn to repeat the 2022 pattern against Lula: a video by city councilman Thomaz Henrique calling him a “thief, criminal, and friend of a narco-dictator” was taken down within 24 hours.

The pattern repeats: concentrated power, a narrow monocratic or panel decision, and a ruler that never appears the same size for both sides. The technical defense offered in cases involving the current president is always the same: presumption of innocence is a constitutional guarantee. That’s true — and insufficient. A principle applied with such consistency to only one side of the spectrum stops functioning as a principle. It starts functioning as a shield. And in Gilmar’s case, not even that defense exists: there is no constitutional guarantee that justifies opening an inquiry against someone for joking about the Court itself.

And the potential harm here is of a different order of magnitude. This is about giving even more power — to institutions already caught using their own tools unevenly, against opponents on one side and in self-defense on the other — to define, in the middle of an election contest, what counts as truth and what can or cannot be broadcast. A State that becomes referee, protected target, and owner of the narrative all at once is not protecting democracy. It is a genuine affront to it.

In the end, the problem isn’t solved by fixing the ruler. Not every ruler should exist in the first place. Justices no one elected decide what is true, what is humor, what can or cannot reach the voter — and call this protecting democracy. It isn’t. It’s the opposite: it’s democracy deciding less and less about itself, while an elite in robes decides more and more. As long as this keeps being treated as institutional routine, Brazil will keep voting with the ruler in the hands of those who answer to no one.

Originally published in Portuguese by Crusoé.

Click here to read the original article.