Brazil’s Superior Electoral Court (TSE) is set to convene its ministers on Tuesday, August 18, 2026, to discuss the use of artificial intelligence in the upcoming general elections and to clarify the scope of rules governing deepfakes in political campaigns.
The meeting, originally scheduled for last Thursday, was postponed after the plenary session ran longer than expected. Court President Kassio Nunes Marques rescheduled the gathering to ensure ministers could deliberate on the matter without time pressure.
According to CNN Brasil, the justices are expected to reaffirm the ban on deepfake material during elections and to align their understanding of which situations should fall under that prohibition. The discussion does not seek a blanket ban on all uses of AI in campaigning. Instead, the aim is to draw a clearer line between synthetic content that is permitted and content that falsely attributes statements, gestures or actions to a person that never occurred.
Under the interpretation under discussion, AI may continue to be used in ways that do not create a false manifestation of a person. For example, a campaign could insert a candidate into a clearly artificial scene, provided all other electoral rules are respected.
The issue arises when the technology is used to make someone appear to say or do something that did not happen. The tendency is that such content will be considered a prohibited deepfake even if the fabricated portrayal favors the candidate or reproduces a position the depicted person would agree with.
The session will serve to harmonize this interpretation among the ministers before the court judges concrete cases already filed with the tribunal.
One case that could be directly affected is a lawsuit brought by the Workers’ Party (PT) against the presidential campaign of Senator Flávio Bolsonaro (PL–RJ). The action challenges a video produced with AI and displayed during the convention that formalized Flávio’s candidacy. In the video, a digital reconstruction of former President Jair Bolsonaro (PL) declares support for his son.
Flávio’s campaign denies wrongdoing, arguing that the audience was informed the video was created with artificial intelligence. Nunes Marques intends to include the case on the court’s docket after the ministers build a shared understanding of the rules’ reach.
If the court concludes that recreating a person to attribute a nonexistent manifestation constitutes a prohibited deepfake, the TSE will still need to determine whether the video used by Flávio Bolsonaro’s campaign fits that definition and what penalties, if any, would apply to the senator.
The ruling is expected to serve as a benchmark for other AI‑related cases that may reach the Electoral Justice during the 2026 campaign.