[REVISION HISTORY]

Brazil 2026 AI Deepfake Controversy Continues

Updated 23 times since CLSTR started tracking revisions of this situation.

What changed

2026-08-23 16:26 UTC → 2026-08-28 20:30 UTC · added ~~removed~~

The legal dispute regarding AI-generated content in Brazilian politics intensified in August 2026. The Federação Brasil da Esperança coalition petitioned the Superior Electoral Court (TSE), arguing that TSE Resolution No. 23.732/2024 imposes an “objective prohibition” on synthetic content that alters a person’s image or voice to benefit or harm a candidacy, regardless of whether there is an intent to deceive. In response, Flávio Bolsonaro’s counsel maintained that deepfakes require a “verified attempt to deceive” and that digital characters and animations are standard, constitutionally legitimate tools in political communication. On 21 August, TSE Minister André Mendonça ordered the removal of a separate deepfake video featuring Senator Flávio Bolsonaro. ~~Published by the profile ‘Brasil Sátira do Poder’, the video used AI to depict the candidate in fabricated scenarios involving the transport of money alongside Daniel Vorcaro.~~ Mendonça ruled that while AI use in elections is not broadly prohibited, it must follow strict transparency rules. He distinguished this content from ~~caricatures or animations,~~ caricatures, noting ~~that~~ the video was designed to “mimic real-life events and behaviors” ~~and~~ to create non-existent scenarios presented as reality. ~~In response~~ By late August, the regulatory focus shifted toward establishing formal definitions. Minister Mendonça proposed a legal framework to ~~these technological challenges,~~ define ‘deep fakes’—described as synthetic audio or video that creates a false perception of authenticity—to provide guidance for future rulings on AI-generated content. This follows ongoing efforts by the Regional Electoral Court in Mato Grosso ~~has invited candidates~~ to ~~participate in a voluntary initiative aimed at protecting~~ protect candidates against ~~AI-driven~~ voice ~~cloning. Concurrently,~~ cloning and the ~~TSE is evaluating~~ TSE’s evaluation of new monitoring mechanisms ~~to monitor free electoral advertising on~~ for radio and television advertising to prevent ~~fraud following complaints regarding missed broadcasts.~~ fraud.

Versions

Only revisions since CLSTR began indexing content versions appear here. Select a version to see what changed compared to the one before it.