Nearly two-thirds of publications (63%) that used artificial intelligence tools during this year's pre-election campaign in Brazil did not disclose the use of the technology, violating the rules of the Superior Electoral Court (TSE). The research also revealed that 81% of identified cases involved the manipulation of images or audio of public figures through deepfake technique, a practice that is banned by Electoral Courts in certain situations.
The survey was conducted by the Observatory AI in Elections, an initiative of the organization Data Privacy Brasil in conjunction with the research laboratory Aláfia Lab. The study analyzed the period from January 1 to August 15, considered the pre-campaign, and identified 413 cases of electoral posts produced with the aid of AI.




