Learn about the rules regarding the use of AI during the 2026 Elections.
Table of Contents
A artificial intelligence will be subject to stricter rules in Elections 2026To combat risks such as misinformation, deepfakes and digital manipulation, the Superior Electoral Court (TSE) Brazil has approved new rules affecting candidates, political parties, digital platforms, and technology companies. The measures define limits on the use of AI during campaigns and create mechanisms to increase the transparency and security of information reaching voters. See the main details below:
Impact of AI on elections and new TSE rules
A artificial intelligence It is already part of the daily lives of millions of Brazilians and, naturally, has also begun to influence how political information is produced, shared, and consumed. Tools capable of creating texts, images, audios, and videos in a few seconds expand the communication possibilities of election campaigns, but also increase the risks of disseminating false, manipulated, or out-of-context content.
Considering this new digital context, the Superior Electoral Court (TSE) approved a series of specific rules for the Elections 2026Seeking to balance technological innovation and the protection of democracy, the measures have been incorporated into the regulation of electoral propaganda and establish clear limits for the use of AI by candidates, political parties, digital platforms, and companies responsible for artificial intelligence systems. The goal is to ensure that voters have access to transparent information and can make their decisions without being misled by deceptive or manipulated content.
Among the main new features are the mandatory identification of content produced or altered by AI, the prohibition of tools that recommend or rank candidates, restrictions on the use of synthetic materials in the days leading up to the vote, and the creation of mechanisms to expedite the removal of content considered illegal.
What can candidates do?
Despite the restrictions imposed by the new regulations, the use of AI in campaigns has not been prohibited. Candidates and parties can use tools to create communication materials, develop advertising pieces, produce content for social media, and even optimize voter engagement strategies. However, all use must follow strict criteria of transparency and respect for electoral rules.
Use permitted with mandatory warning.
The main rule established by TSE The requirement is that electoral content produced or significantly altered by artificial intelligence can be used during the campaign, provided there is clear identification of this use. This requirement applies to different formats, including texts, images, videos, and audios, regardless of the platform on which they are disseminated. The intention is to allow voters to know exactly when they are viewing material created or modified with the aid of generative technology. This notice must be presented explicitly, visibly, and in an easily understandable way, without hiding or hindering the information.
Identification of the tool used
In addition to informing users that certain content was created or altered by artificial intelligence, the rules approved by TSE They also require that the technology used in the production of that material be indicated. This means that election campaigns must disclose which AI tool, system, or platform was used to create or modify the content disseminated to the public.
What is prohibited
Although the Superior Electoral Court While the regulation has allowed the use of artificial intelligence in the 2026 election campaigns, it has also established a series of prohibitions to prevent manipulations that could compromise the voter's free choice. These restrictions include the creation of false or misleading content, paid dissemination by third parties, and the use of synthetic materials during critical periods of the election. The goal is to prevent AI technologies from being used to spread disinformation or generate undue advantages for candidates and parties.
Deepfakes and deceptive manipulations
One of the biggest concerns of electoral justice It is the advancement of what are called deepfakesArtificial intelligence-generated content capable of reproducing a person's voice, image, and even gestures with a high degree of realism. The technology allows for the creation of fake videos of candidates making statements that never happened or participating in fictitious situations. In a digital environment marked by the speed of information, this type of content can reach millions of people before it is even debunked, making the fight against such content a difficult one. deepfakes one of the priorities of the electoral rules for 2026.
The risk is not merely theoretical. In recent years, various countries have recorded cases of digital manipulation involving politicians and public figures. In Brazil, experts have warned... Superior Electoral Court about scenarios involving fake videos, compromising montages, and even sexually explicit content generated by AI to attack candidates.
AI-powered content 72 hours before the election
Among the strictest measures approved for the 2026 Elections is the prohibition of publishing, republishing, and boosting new content produced or altered by artificial intelligence during the elections. 72 hours before the vote and also in 24 hours after the closing of the pollsThis rule applies especially to materials that use images, voices, or statements from candidates and public figures, even when the content is correctly identified as being generated by AI.
The justification for TSE It's about avoiding what's called "last-minute bombshell effect"This occurs when false or manipulated information is released in the final moments of a campaign, leaving little time for rebuttal, technical analysis, or response from the affected candidates," according to the minister. Cassio Nunes MarquesThe measure seeks to prevent "unwanted surprises" precisely during the most sensitive period of the electoral process, when viral information can influence the decision of thousands or even millions of voters.
"Cutting Championship" banned
Another practice that came under scrutiny from the Electoral Court was the so-called "cutting championship", a strategy that gained notoriety in the municipal elections of São Paulo in 2024The model consisted of encouraging supporters to produce and share short campaign videos on social media in exchange for rewards, prizes, economic advantages, or participation in performance rankings. This dynamic transformed supporters into a kind of informal political outreach network, significantly expanding the reach of electoral messages.
One of the best-known cases involved the then-candidate for mayor of São Paulo, Pablo Marcal, whose campaign used incentive mechanisms to stimulate the mass production of content by third parties. The repercussions generated debates about abuse of economic power and misuse of digital platforms. In response, the Superior Electoral Court It prohibited any form of contracting, rewarding, or ranking that offers direct or indirect financial benefits to individuals or companies for publishing political-electoral content on their profiles. The measure seeks to prevent the creation of "digital armies" capable of distorting the organic reach of campaigns.
Payment for political promotion
The new rules also prohibit candidates, parties, or third parties from paying influencers, content creators, or ordinary users to publish political campaign advertising on their own profiles. This prohibition applies regardless of the form of remuneration, encompassing money, gifts, prizes, bonuses, indirect benefits, or any other economic advantage offered in exchange for disseminating the content. The measure aims to increase the transparency of public debate and prevent voters from being exposed to political messages without knowing that there was a commercial relationship behind that publication.
Rules for social networks and platforms
The new electoral rules for 2026 don't focus solely on candidates and political parties. For the first time, the TSE established broader obligations for social mediaDigital platforms and companies responsible for artificial intelligence systems. The intention is to create a safer digital environment during the election period, reducing the circulation of misinformation, manipulated content, and practices that could compromise the integrity of the democratic process.
Removing false content
One of the main determinations of TSE Digital platforms must act quickly when they identify illicit content related to elections. This includes posts containing demonstrably false information, materials manipulated by artificial intelligence, content that baselessly attacks the electronic voting system, or messages capable of generating misinformation about the electoral process. This obligation applies to both organically published content and material boosted through paid advertising.
Deletion of fake profiles and bots.
As Elections 2026 They also introduce stricter rules against fake accounts used to manipulate political debates. Proven fictitious, anonymous profiles, or those operated by bots, may be removed from platforms when used to spread electoral disinformation, attack democratic institutions, or disseminate illicit content related to the electoral process. The measure seeks to combat organized structures that use thousands of accounts to artificially amplify certain narratives.
The topic gained particular prominence after several investigations in recent years identified coordinated networks of automated profiles operating in political discussions. With the new rules, platforms now have a duty to act to stop this type of activity when there is evidence that the profiles are being used to undermine the integrity of elections or spread fraudulent information.
Transparency in sponsored posts
Another important change involves the content boosted during election campaigns. TSE It was determined that publications sponsored by candidates and parties must be clearly and unequivocally identified, allowing users to know when they are viewing paid political advertising. This requirement applies to ads displayed on social media, search engines, and other digital platforms used by campaigns.
Platform compliance plans
Companies responsible for social networks, search engines, and artificial intelligence systems will also need to develop what are called... election compliance plansThese documents serve as a kind of prevention and response protocol, detailing what measures will be taken to reduce risks related to disinformation, manipulated content, and other illicit practices during the election period.
The plans should include information on content monitoring, internal moderation processes, communication channels with electoral authorities, and procedures for handling complaints.
The publisher may need to prove its authenticity.
One of the most debated changes to the regulation is the possibility of Reversal of the burden of proof In cases involving content produced or manipulated by artificial intelligence, and in situations where technical proof of fraud is excessively complex for the complainant, the Electoral Court may order the person responsible for the publication to provide evidence of the authenticity of the disseminated material.
This means that a person, campaign, or group accused of disseminating manipulated content may be required to technically demonstrate that the material is legitimate or explain in detail how it was produced. The measure was created because the evolution of AI technologies has made... It is becoming increasingly difficult to identify digital forgeries. Based solely on visual analysis.
What AI agents can and cannot do
With the advancement of generative artificial intelligence tools, millions of people have started using chatbots and virtual assistants to search for information about politics, candidates, and elections. Given this scenario, the... TSE created specific rules to define how these platforms can operate during the election period. The intention is to allow systems such as ChatGPT, Gemini:, Claude, Perplexity, AI goal and others continue to provide useful information to citizens, but without exerting undue influence on voting decisions or favoring specific candidates and parties.
To provide information about candidates and elections.
Artificial intelligence platforms remain authorized to provide information about candidates, political parties, contested positions, the electoral calendar, and the workings of the democratic process. A voter can ask who the candidates for President of the Republic are, what the responsibilities of a senator are, or how the Brazilian electoral system works, and the AI can respond normally, as long as it presents factual and impartial information.
Furthermore, the systems can explain government proposals, present public biographical data, contextualize political events, and summarize news of electoral interest. The goal is for these tools to function as instruments of information and civic education.
Answering voters' questions
Another permitted use involves clarifying doubts related to the exercise of citizenship. AI agents can guide voters on election deadlines, voter registration, polling places, electoral justifications, campaign rules, and the functioning of electronic voting machines. This capability can broaden the population's access to relevant information, especially for people who have difficulty finding official data or understanding legal and electoral terms.
It is forbidden to recommend candidates or parties.
The main restriction imposed on artificial intelligence systems involves electoral recommendations. Platforms are prohibited from suggesting, ranking, prioritizing, or indicating candidates and political parties, even when the user asks direct questions such as "Who should I vote for?" or "Who is the best candidate for president?". The Electoral Court's concern is to prevent algorithms from influencing decisions that should be made exclusively by citizens.
The relevance of this rule became evident after tests conducted by BBC News BrazilThe analyses showed that different AI systems exhibited distinct behaviors when answering election questions. While some platforms avoided direct recommendations, others went so far as to present rankings, assign positive or negative characteristics to certain politicians, and even suggest specific names in some situations.
It is forbidden to create content that promotes political or sexual violence against candidates.
The new rules also prohibit the use of artificial intelligence to create content that promotes political violence or exposes candidates to degrading, humiliating, or offensive situations. The prohibition includes fake images, manipulated videos, doctored images, and any other AI-generated material aimed at attacking the reputation of electoral participants through fraudulent information or abusive content.
One of the focuses of the regulation is combating gender-based political violenceThis is a problem that has worried electoral authorities in several countries. Therefore, AI systems are prohibited from generating images, videos, or audios with sexual content involving candidates, especially when the objective is to embarrass, intimidate, or harm the political participation of women. The measure was adopted in response to the global increase in cases of... pornographic deepfakes, technology used to digitally insert people's faces into fake intimate content.
How will the Justice system oversee this?
To monitor the use of artificial intelligence During an election, the use of AI presents an unprecedented challenge for the Electoral Court. Unlike traditional advertising, AI-generated content can be produced on a large scale, customized for different audiences, and disseminated in minutes through social networks, messaging apps, and digital platforms. To address this scenario, the... Superior Electoral Court (TSE) It expanded monitoring mechanisms and created rules that involve not only candidates and parties, but also technology companies, digital platforms, and developers of artificial intelligence systems.
One of the main strategies will be joint action between the Electoral Court and institutions specializing in technology. The new rules allow electoral courts to establish partnerships with universities, research centers, and technical bodies to analyze suspicious content. These specialists will be able to assist in identifying... deepfakesThis includes digital montages, synthesized audio, and other materials generated or altered by artificial intelligence. The goal is to provide technical support so that judges and courts can more accurately assess the authenticity of content circulating during the election campaign.
In addition to institutional oversight, the process will also involve the participation of society and the digital platforms themselves. Tools such as... Electoral Disinformation Alert System (SIADE) They allow citizens to report content that is considered false, misleading, or out of context.
And you? Were you already aware of all these rules regarding the use of AI during the Elections 2026Do you agree or disagree with any of these statements? Tell us in the comments below.
See also:
Sources: Senate Verifies, TSE e Data Privacy Brazil Research.
Discover more about Showmetech
Sign up to receive our latest news via email.
