Brazilian Superior Electoral Court (TSE) Establishes Legal Standard on Deepfakes and Launches Partnership with Google to Combat the Misuse of AI in the 2026 Elections

Brazilian Superior Electoral Court (TSE) Establishes Legal Standard on Deepfakes and Launches Partnership with Google to Combat the Misuse of AI in the 2026 Elections

On August 26, the Brrazilian Superior Electoral Court (TSE) and Google Brazil launched a campaign to enhance the protection of candidates in the 2026 General Elections against the misuse of artificial intelligence (AI). A few days later, on September 1, the TSE Plenary, by majority vote, established criteria for the characterization of deepfakes.

The legal holding was established in the judgment of Case No. 0601315-97.2026.6.00.0000, which addressed the application of Article 9-C, § 1, of TSE Resolution No. 23,610/2019. Among other restrictions, the provision prohibits the use of deepfakes to harm or favor a candidacy in electoral propaganda. According to the TSE, for the purposes of this provision, the characterization of a deepfake requires “synthetic content produced or manipulated through artificial intelligence or equivalent technology, possessing a degree of realism or verisimilitude and intended to create, reproduce, or alter the image, voice, or expression of a living, deceased, or fictional person.” The Court further clarified that the prohibition set forth in Article 9-C, § 1, applies when the content is characterized as electoral propaganda.

In parallel, the campaign launched by the TSE and Google seeks to mitigate the risks associated with the improper use of deepfakes in the 2026 General Elections by encouraging candidates to use YouTube’s Likeness Detection tool (Likeness ID). To use the tool, candidates must have a YouTube account, access the feature through YouTube Studio, accept the applicable terms of use, and complete an identity verification process by submitting an official identification document and a selfie in both photo and video formats.

Following registration and processing of the submitted information, YouTube creates a reference representation of the individual’s appearance to assist in identifying videos that reproduce their likeness. The platform then conducts checks on published content and notifies the user when it identifies potential matches. Based on these notifications, candidates may review the identified videos and, if they determine that their likeness has been used improperly, request removal of the content. Removal is subject to the platform’s review and may take place where the content violates YouTube’s policies.

Taken together, the legal holding concerning deepfakes and the campaign launched in partnership with Google demonstrate complementary approaches to addressing the risks arising from the improper use of AI and the reproduction of individuals’ likenesses during the electoral period.

The ruling can be accessed through the following link: Ruling

Note: For quick release, this English version is provided by automated translation without human review.