Electoral Court approves restrictions on AI use in 2026 elections
Brazil’s Superior Electoral Court approved on Monday (Mar. 2) rules governing the use of artificial intelligence (AI) during the general elections in October this year. The rules apply to candidates and political parties.
The court unanimously banned the posting of modified content on social media in the 72 hours before the election and 24 hours after the vote.
The restriction applies to modifications involving the images and voices of candidates or public figures. The first round is scheduled for October 4.
The rules were established through the approval of several resolutions that will govern the election, in which voters will choose the president, governors, senators, and federal, state, and district representatives.
The court also prohibited AI providers from offering, even at users’ request, suggestions of candidates to vote for. The aim is to prevent algorithmic interference in voters’ free choice.
To combat digital misogyny, the Superior Electoral Court prohibited social media posts featuring montages of female candidates, as well as photos and videos containing nudity or pornography.
The Electoral Court also reaffirmed that internet providers may be held accountable if they fail to remove fake profiles and illegal posts created by their users.
Freedom of expression
To prevent episodes of censorship seen in previous elections, the justices decided to allow spontaneous political expressions during the pre-campaign period in universities, schools, and social movement spaces.
Candidates are also allowed to distribute leaflets in streets, parks, and squares, provided that the free movement of people in these areas is maintained.
Last week, the court approved seven additional resolutions for the elections. The texts cover the dissemination of electoral polls, voter transportation, fundraising, accountability, and changes to the voter registration system.
