Quick Answer:
Brazil's Superior Electoral Court received 87 AI-related complaints between January and May 2026, four times the rate seen in the same period of 2022, according to The Rio Times. Brazil approved one of the world's strictest AI election regulations in March 2026, but experts warn of real enforcement gaps ahead of the October vote.
Key Takeaways:
- Complaint surge: 87 formal AI complaints were filed with the TSE between January and May 2026, nearly four times the same period in 2022, according to The Rio Times.
- New rules: the TSE banned deepfakes outright, mandated labeling of AI-generated content, and blocked all AI content 72 hours before voting and 24 hours after, per The Rio Times.
- Maximum penalty: a candidate using deepfakes risks annulment of their candidacy registration or mandate, The Rio Times reports.
- Identified gap: electoral law professor Fernando Neisser told Tech Policy Press that if even the Federal Police cannot reliably identify a deepfake, it is unrealistic to expect platforms to detect it consistently.
- Gray zone: researcher Erick Beyruth told Tech Policy Press that the current rule text creates the impression that WhatsApp and Telegram fall outside the reach of electoral rules.
Brazil heads into its October 2026 presidential election with more than 155 million eligible voters and one of the most detailed AI regulatory frameworks in the world applied directly to the electoral process. At the same time, it faces the biggest real-world stress test that framework has seen so far: a measurable surge in AI-content complaints targeting the leading candidates, months before polls even open.
For businesses across the U.S. and Latin America that already depend on AI representing them accurately to customers, Brazil's case is an early warning. If a court with a dedicated budget, specialized lawyers, and rules on the books since 2024 still struggles to separate real content from manipulated content, no business should assume by default that AI describes it accurately.
The Jump in Complaints: 2022 to 2026
According to The Rio Times, between January and May 2026 Brazil's Superior Electoral Court received 87 formal complaints tied to the leading presidential contenders, roughly four times the number recorded in the same period of the 2022 election cycle. More than half of those complaints trace back to two fronts: President Lula's Workers' Party and Senator Flávio Bolsonaro's Liberal Party, the same source reports.
The Rio Times reports that Lula has been the target of AI-generated videos falsely linking him to financial scandals, failed-bank issues, and pension fraud. On the other side, Senator Flávio Bolsonaro has been accused of using AI to tie him to that same failed-bank affair, while his party has also filed complaints against the opposition. The outlet also documents the use of AI-generated virtual personas designed to impersonate real people and spread misinformation, as well as a dispute over a samba school parade themed around the president's political history.
Brazil's New AI Election Rules, Explained
Per The Rio Times, the Superior Electoral Court approved a set of rules in early March 2026 that expands the framework first introduced for the 2024 municipal elections. The core provisions include:
What Brazil's new AI electoral rules prohibit:
- Deepfakes banned outright from electoral propaganda
- Full blackout of AI-altered content on social media 72 hours before voting through 24 hours after polls close
- All AI-generated campaign material must carry a clear, prominent, and accessible disclosure label
- AI systems are prohibited from ranking or recommending candidates
- Specific safeguards against AI-generated sexual or pornographic deepfakes targeting candidates
The penalties are designed to be severe. The Rio Times reports that a candidate using deepfakes risks annulment of their candidacy registration or mandate. Platforms, meanwhile, face joint civil and administrative liability if they fail to remove illegal content immediately after being notified. The framework also introduces a procedural innovation: in AI-related election lawsuits, the burden of proof can reverse, meaning a publisher may need to prove content authenticity rather than an accuser proving manipulation.
Tech Policy Press describes the framework itself as one of the world's most detailed regulatory frameworks for artificial intelligence in elections, and campaign strategists cited by that outlet expect 2026 to be remembered as the year of AI in Brazilian politics.
The Gaps That Worry Experts
Despite the regulation's scope, Tech Policy Press documents serious warnings from specialists about concrete enforcement gaps. Fernando Neisser, an electoral law professor at FGV São Paulo, framed the underlying problem directly: "If the user can't identify it, the courts can't identify it, and even the Federal Police struggle — how can we expect Instagram to detect it?" According to Neisser, the current rules concentrate detection responsibility on platforms rather than on the developers who build the AI tools.
Erick Beyruth, a researcher at PUC São Paulo, warned Tech Policy Press about a second gap: the rule text, as written, "creates the impression that these environments are also immune to electoral rules," referring to private messaging apps like WhatsApp and Telegram — a significant blind spot in a country where those apps are a mass channel for political distribution. Electoral attorney Alberto Rollo added that identifying AI manipulation with certainty remains technically complex even with forensic analysis, according to Tech Policy Press.
The problem in one sentence: Brazil wrote detailed rules for a phenomenon that its own forensic experts cannot detect with absolute certainty, and placed the enforcement burden on platforms rather than on the developers who build the technology.
The Dona Maria Case: What the Threat Looks Like in Practice
Tech Policy Press documents a concrete case that illustrates why this regulation arrived under pressure. An Uber driver from outside Rio de Janeiro created "Dona Maria," an AI-generated persona that, per the outlet, reached one video with 8 million views and more than 20,000 comments, with twelve separate videos each surpassing 1 million views. In late April 2026, three coalition parties filed a lawsuit against "Dona Maria" in electoral court, according to Tech Policy Press. The case shows the phenomenon is not limited to official party campaigns: independently produced, low-cost content can reach mass audiences before any verification mechanism catches up.
What This Means for Your Business
Brazil's case is not only a story about electoral politics. It is the most visible proof yet of a problem already affecting any business with a digital presence: AI tools generate and distribute content about people and brands faster than any verification system, human or automated, can confirm it. If Brazil — with a dedicated court, a budget, and rules on the books since 2024 — still faces a fourfold increase in complaints, the average business has no equivalent mechanism protecting it from ChatGPT, Perplexity, or Google AI Overviews describing it incompletely, out of date, or simply wrong.
This is exactly where MerchandisePROS's AI Search Optimization (AEO) service comes in. The question every business should be asking is not whether AI will generate content about its brand, but whether that content will be accurate when a potential customer asks ChatGPT or Perplexity who offers your service in your city. Our free audit checks exactly how the leading AI tools represent your business today and delivers a 0-to-100 score with a prioritized action plan, in under 60 seconds.
Frequently Asked Questions
How many AI-related complaints has Brazil's electoral court received in 2026?
Between January and May 2026, Brazil's Superior Electoral Court (TSE) received 87 formal complaints involving the leading presidential contenders, roughly four times the number filed during the same period of the 2022 election cycle, according to The Rio Times.
What do Brazil's new AI electoral rules prohibit?
Rules approved by the TSE in early March 2026 ban deepfakes outright in electoral propaganda, require clear, prominent and accessible disclosure labels on all AI-generated campaign material, prohibit AI systems from ranking or recommending candidates, and impose a full AI-content blackout 72 hours before voting through 24 hours after polls close.
What happens if a candidate uses a deepfake in their campaign?
According to The Rio Times, candidates who use deepfakes risk annulment of their candidacy registration or mandate. Platforms face joint civil and administrative liability if they fail to remove illegal content immediately after being notified.
What gaps exist in Brazil's AI election regulation?
Experts cited by Tech Policy Press say the rules concentrate enforcement responsibility on platforms rather than AI developers, leave private messaging apps like WhatsApp and Telegram in a legal gray zone, and face the underlying problem that identifying AI manipulation with certainty remains technically complex even with forensic analysis.
What does this mean for businesses outside of politics?
If Brazil's electoral courts, with detailed rules and dedicated budget, still struggle to verify which AI content is real, an average business cannot assume AI represents it accurately by default. The way to find out is to audit how AI tools describe you today.
"If Brazil, with a dedicated electoral court and rules on the books since 2024, still can't verify all the AI content about its own candidates, no business should assume AI describes it correctly without checking first."
- Diego Medina F, Founder of MerchandisePROS