The use of AI () has grown in political communication and is expected to intensify in 2026. Parties and candidates are already using it to attack opponents and spread ideas, while the (Superior Electoral Court) rules adopted in the last municipal election.
Last Monday (19), the court released the initial proposal for the rules on electoral advertising for this dispute without foreseeing changes to the items on the topic. A public consultation period was opened, with hearings scheduled for February.
For example, the use of calls and “fabricated or manipulated” content with the aim of disseminating untrue or seriously decontextualized facts is already prohibited. For cases that do not fall within prohibited conduct, it is mandatory that the use of AI in the content be informed.
In 2024, there were fears that the use of technology could unbalance the elections in a widespread way, which has not been confirmed, according to experts. The assessment now is that the challenge tends to intensify with the expanded use of these tools in the next dispute.
In recent months, for example, PT members have flooded platforms with pieces that exploit . The strategy, which responded with videos also produced with the same technology.
The actions cross the political spectrum and mix humor, political attack and agenda defense. The pieces range from images inspired by popular styles on social media to the generation of realistic videos.
“The arrival of generative AI allowed campaigns to create thousands of personalized messages in seconds, something unthinkable a decade ago”, says Jaqueline Zulini, doctor and professor of political science at FGV (Fundação Getulio Vargas).
For her, technology has expanded access to campaign tools, even if it brings new risks to electoral integrity.
“What used to be the privilege of large campaigns is now within the reach of any candidate,” he says.
Professor of political communication and marketing at ESPM (Escola Superior de Propaganda e Marketing) Marcelo Vittorino states that use is still below potential. “Most parties are in the shallow end of the pool,” he says. According to him, the main effect so far is the reduction of costs and the increase in the volume of content produced.
In the opinion of Pedro Burgos, AI consultant and communication professor at Insper (Institute of Education and Research), the main recent change was the migration from text to multimedia content. “Before, AI helped to write speeches. Today it is much better for creating images, videos and audio”, he says.
In addition to content production, the professor cites expanding uses of AI. One possibility is to improve “social listening”, a practice that monitors conversations on social networks to map themes and demands from the electorate.
“For a long time this was a very inexact science. Today we can do more complex things, trying to detect which subjects are most followed, most seen and most important to the population”, he says.
Burgos assesses that the main risk related to disinformation is not just realistic deepfakes. He points out simpler and more frequent manipulations, such as true videos taken out of context.
“Speech manipulated by editing or with altered speed, for example, tends to be more common than ‘perfect’ deepfakes all the time.”
Parties adopt a cautious tone when dealing with the issue. The states that it does not yet have formalized guidelines for the use of artificial intelligence. According to the caption, communications teams are instructed to follow “the democratic values of transparency, responsibility and the use of true data and facts.” The party also says that all content produced with AI must undergo human supervision.
The PP states that the guidance for candidates is to observe the TSE resolutions. “For us, the fundamental thing is to combat disinformation and respect voters’ right to an honest electoral process”, says the caption.
The says that technology should be used in an auxiliary way and that strategies for 2026 are still under construction. According to the party, the trend is to employ artificial intelligence as a “technical support tool”, in areas such as data analysis, information organization, network monitoring and support for content production. The acronym says that use should not “replace human action or political responsibility for decisions”.
The MDB states that it will follow “respect for electoral legislation and an ethical commitment to the truth”. The party says it will use the technology within these precepts and did not detail specific applications.
He says he is not opposed to the use of artificial intelligence and considers technology an ally of creativity, innovation and efficiency in political communication.
The caption states that it does not use artificial intelligence to “create fictitious people, simulate demonstrations, fabricate images of real events or produce content that represents facts as if they were authentic records.” The PT also says that, whenever technology is used, this is explicitly indicated to the public.
In relation to other parties, the party claims to view the scenario with concern. According to the PT, there are risks of using AI for disinformation, image manipulation, omission of contexts and dissemination of fake news, practices it says it rejects.
