They had acted as though the technology was theirs to use
In the months before Brazil's 2026 presidential race, a political party reached beyond the boundaries of consent and law, using artificial intelligence to fabricate a video of a former Supreme Court justice without his knowledge. The Christian Democracy party's deployment of a deepfake likeness of Joaquim Barbosa — a man of institutional gravity and electoral ambition — exposed not only the party's internal disarray but a wider vulnerability in democratic life: that technology now permits the convincing forgery of human speech and presence, and that the rules governing such forgery have not yet caught up. This moment arrived as a test, asking whether Brazilian democracy would treat synthetic political media as a serious threat or an acceptable convenience.
- A retired Supreme Court justice discovered his face and voice had been fabricated by a political party he never authorized to represent him.
- The deepfake surfaced amid open warfare inside the Christian Democracy party, including the expulsion of veteran figure Aldo Rebelo over competing claims to the presidential nomination.
- The incident forced an urgent question into Brazilian public life: if a major party can deploy AI-generated likenesses without consequence, what stops the practice from spreading unchecked?
- Major outlets including Folha de S.Paulo and CartaCapital began treating the case not as a party scandal but as a referendum on electoral integrity in the age of synthetic media.
- Regulators and observers now face pressure to determine whether existing electoral law can reach this new form of political fabrication — or whether new frameworks must be built from scratch.
In May 2026, Brazil's Christian Democracy party posted a video of Joaquim Barbosa — a retired Supreme Court justice who had announced presidential ambitions — without his knowledge or consent. The video was a deepfake, constructed through AI to simulate his appearance and speech as though he had recorded it himself.
Barbosa had authorized nothing. The act raised immediate questions about where political campaigning ends and fabrication begins, particularly as technology grows capable of placing words in anyone's mouth with convincing fidelity.
The incident did not occur in a vacuum. The party was already tearing itself apart over its presidential nomination. Aldo Rebelo, a longtime party figure, had been expelled amid the dispute, and competing factions were maneuvering publicly and bitterly. The deepfake arrived as a symptom of how far the party was willing to reach in shaping its own narrative.
Beyond the internal chaos, the episode pointed toward something larger: a new frontier in political manipulation, where synthetic media can be weaponized against the very figures it claims to represent. Observers and journalists began asking what safeguards existed, and whether Brazilian electoral law had any meaningful answer to AI-generated campaign content.
As the story moved through major Brazilian outlets, it became a test case rather than merely a scandal. Barbosa's institutional stature gave the incident weight that a lesser figure might not have commanded. The question left hanging was whether regulators would act — and whether this moment would force Brazil to build new rules before the next fabrication arrived.
In May 2026, Brazil's Christian Democracy party found itself at the center of a controversy that exposed both the party's internal fractures and the emerging risks of artificial intelligence in electoral politics. The party's social media profile had posted a video featuring Joaquim Barbosa, a retired justice of Brazil's Supreme Court, without his knowledge or consent. The video was not authentic—it had been generated using AI technology, a deepfake constructed to appear as though Barbosa himself had recorded it.
Barbosa, a prominent jurist who had announced his intention to run for president, had not authorized the party to create or distribute the video. The unauthorized use of his likeness raised immediate questions about the boundaries of political campaigning in an age when technology can convincingly fabricate speech and appearance. The Christian Democracy party, in posting the content, had crossed a line that Brazilian electoral law and basic principles of consent had not yet fully anticipated.
The timing of the incident was not incidental. The party was convulsing over its own presidential nomination. Aldo Rebelo, a longtime party figure, had been expelled from the Christian Democracy amid the dispute over who would carry the party's banner into the election. The internal warfare had become public and bitter. Flávio, another party actor, had begun signaling toward Rebelo even as the party leadership moved to remove him from contention. The video incident arrived in the midst of this chaos—a symptom, perhaps, of how far the party was willing to go to shape its political narrative.
The use of an AI-generated video to represent a public figure without authorization raised stakes beyond the immediate party dispute. It suggested a new frontier in political manipulation: the ability to put words in a candidate's mouth, to create false evidence of statements never made, to weaponize technology against the very people it purported to promote. Barbosa had not consented to being the subject of this fabrication, and his silence on the matter—or his objections—would likely shape how Brazilian regulators and the public understood the boundaries of acceptable campaign practice.
The incident prompted reflection among observers and journalists about electoral integrity in Brazil. If a major party could deploy deepfake technology without consequence, what would prevent others from doing the same? What safeguards existed? The Christian Democracy's actions suggested that the legal and ethical frameworks governing political speech had not kept pace with the capabilities of artificial intelligence. The party had acted as though the technology was theirs to use, the image of a public figure theirs to manipulate, without needing permission or bearing responsibility for the fabrication.
As the story circulated through Brazilian media outlets—Poder360, Folha de S.Paulo, CartaCapital, and others—it became clear that this was not merely a party scandal but a test case for how Brazilian democracy would reckon with AI-generated political content. Barbosa's status as a former Supreme Court justice gave the incident additional weight; it was not a marginal figure being deepfaked, but a man of institutional standing and electoral ambition. The question now was whether regulators would act, whether the party would face consequences, and whether this moment would catalyze new rules around the creation and distribution of synthetic media in campaigns.
Citações Notáveis
The video was not authentic—it had been generated using AI technology, a deepfake constructed to appear as though Barbosa himself had recorded it.— reporting on the incident
A Conversa do Hearth Outra perspectiva sobre a história
Why would a political party risk this? Using someone's face without permission seems obviously reckless.
It does now. But in the moment, they may have calculated that the video would amplify their message faster than any traditional statement could, and that Barbosa—a potential rival—wouldn't want the legal fight that contesting it would bring.
So they were betting on his silence?
Or betting that by the time he responded, the video would have already done its work in the minds of voters and party members. The party was fracturing over the nomination anyway. The video was a move in that internal war.
What does it say about Brazilian politics that this happened at all?
That the technology arrived faster than the guardrails. No one had written the rules yet. The party acted as though AI-generated content was just another tool, like a edited photograph or a voiceover. They hadn't reckoned with the fact that deepfakes feel different to people—they feel like a violation in a way a traditional ad doesn't.
Will this change anything?
It might. Once a major party crosses a line this visible, regulators usually move. The question is whether they move fast enough to matter in the next election cycle.