The machines work. The information around them is what we need to guard.
In Brazil, a new steward of democratic machinery has taken his post at a moment when the tools of elections are trusted by millions yet questioned by many. Nunes Marques assumed the presidency of the Superior Electoral Court not merely to administer votes, but to defend the very idea that they can be counted fairly. His arrival marks a turning point in how a major democracy confronts not only old doubts about its voting infrastructure, but an emerging and more elusive adversary: the artificial reshaping of truth itself.
- Brazil's electronic voting system — decades old and deeply embedded in national identity — faces renewed scrutiny at the precise moment its new guardian takes office.
- Artificial intelligence has emerged as the court's most urgent concern, not for its ability to hack machines, but for its power to poison the information environment before a single vote is cast.
- Marques drew a deliberate line between what is working — the urns, the audits, the infrastructure — and what is newly vulnerable: the narratives, deepfakes, and disinformation that no ballot box can stop.
- The TSE is signaling a shift from purely technical defense toward a broader battle for public confidence in an era when trust itself has become the primary target.
- Whether institutional credibility can be sustained depends on transparency, cross-institutional coordination, and the agility to adapt as AI-driven threats continue to evolve ahead of future elections.
When Nunes Marques assumed the presidency of Brazil's Superior Electoral Court, he inherited both a legacy and a crisis of confidence. The electronic voting system that has quietly anchored Brazilian democracy for decades was under scrutiny — and rather than deflect, Marques chose to defend it directly, arguing that the urns remain secure and that the electoral process, at its core, continues to function.
But his more striking message concerned what lies beyond the machines. Artificial intelligence, he warned, represents a new category of threat — not one aimed at miscounting votes, but at distorting the information environment that surrounds an election. Deepfakes, coordinated disinformation, and algorithmically amplified doubt can erode public trust long before voters reach the ballot, and no voting machine, however well-audited, can defend against that alone.
The electronic urns have processed millions of votes across many election cycles, surviving technical scrutiny and political skepticism alike. Their resilience is real — but so is their fragility. Confidence in democratic systems is not self-sustaining; it must be actively maintained, especially when the tools of manipulation grow more sophisticated by the year.
Marques now faces the harder task: translating institutional credibility into lasting public faith. That means not only preserving the integrity of the voting infrastructure, but building new defenses against threats that are diffuse, fast-moving, and deeply embedded in the media ecosystems through which most citizens experience democracy. The machines may be secure. The battle for the minds of voters is only beginning.
Nunes Marques stepped into the presidency of Brazil's Superior Electoral Court on a day when the institution faced questions it had not confronted before. The electronic voting system that has anchored Brazilian elections for decades—a source of national pride and technical sophistication—was under scrutiny. But Marques did not retreat into defensiveness. Instead, he mounted a direct case for the machines themselves, arguing that the urns remain secure and that Brazil's electoral process, by and large, functions as it should.
The timing of his assumption of leadership placed him at the center of a conversation that has grown louder across democracies worldwide. Artificial intelligence, he warned, represents a new and distinct threat to electoral integrity. Not the voting machines themselves, but the tools that could be deployed to manipulate information, seed doubt, and distort the democratic process before voters ever reach the ballot. It was a careful distinction—one that acknowledged legitimate technological anxiety while defending the specific infrastructure Brazil had built.
Marques inherited an institution tasked with an enormous responsibility: maintaining public faith in the machinery of democracy at a moment when that faith faces pressure from multiple directions. The electronic voting system has been in place long enough to have become almost invisible to most Brazilians—it simply works, year after year, election after election. But invisibility is fragile. A single major failure, or sustained questions about security, could shatter the confidence that allows millions of people to accept the outcome of an election they did not personally witness being counted.
The focus on artificial intelligence signals a shift in how electoral authorities think about threats. The danger is no longer primarily about the machines being hacked or votes being miscounted in the traditional sense. Instead, it is about the information ecosystem surrounding the election—the narratives that shape how people understand what is at stake, the deepfakes that could undermine trust in candidates or institutions, the coordinated disinformation campaigns that could exploit divisions. These are threats that no voting machine, however secure, can defend against alone.
Marques' position reflects a broader institutional maturity. Rather than dismissing concerns as unfounded or retreating into technical jargon, he acknowledged the real challenge while making a case for what is actually working. The electronic urns have processed millions of votes across multiple election cycles. They have been tested, audited, and subjected to the kind of scrutiny that comes with being a target of skepticism. The system is not perfect—no system is—but it has proven resilient.
What remains to be seen is whether Marques can translate this defense into sustained public confidence. The work ahead involves not just maintaining the technical integrity of the voting infrastructure, but also building defenses against the newer, more diffuse threats that artificial intelligence enables. It requires coordination with other institutions, transparency about vulnerabilities and how they are being addressed, and a willingness to adapt as the threat landscape evolves. The electronic voting system may be secure, but the broader electoral process exists within an information environment that is becoming harder to control and easier to manipulate.
Notable Quotes
Marques defended the electronic voting urns as secure while warning that artificial intelligence poses a distinct and growing risk to electoral integrity— TSE leadership position
The Hearth Conversation Another angle on the story
Why does a new TSE president need to defend the voting machines themselves? Aren't they already proven?
They are proven in a technical sense—they work, they count votes accurately. But proof is not the same as trust. When people start asking questions, silence looks like evasion. Marques is saying: yes, we know you're worried, and here's why you shouldn't be.
But he's also warning about artificial intelligence. That seems like a different problem entirely.
Exactly. The machines themselves aren't the vulnerability anymore. The vulnerability is everything around them—the information people consume before they vote, the narratives they believe about what's at stake. AI makes that landscape much harder to navigate.
So the electronic urns are actually the most trustworthy part of the system?
In a way, yes. They're physical, they're auditable, they leave a paper trail. What you can't audit is whether someone used a deepfake to convince voters that a candidate said something they didn't, or whether a coordinated bot network shaped what people saw in their feeds.
That sounds like a problem that goes beyond electoral authorities.
It does. But the TSE has to work with what it controls and be honest about what it doesn't. Marques seems to be doing both.