Brazil's TSE to judge Bolsonaro avatar case, may set deepfake precedent

Brazil's Superior Electoral Court (TSE) is set to hear a significant case on Tuesday, August 1st, concerning the use of an AI-generated avatar of former President Jair Bolsonaro during a convention for the Liberal Party (PL). The hearing is expected to result in a majority ruling against the tool's use, particularly as it was employed by Senator Flávio Bolsonaro, who is running for president.

The lawsuit was filed by Lula's (PT) coalition, which argues that the use of a video featuring an "AI Bolsonaro" asking for votes violates electoral norms. This is especially relevant as the real Jair Bolsonaro is currently imprisoned. The TSE's decision could set a crucial precedent for how deepfakes and other AI-generated content are regulated in Brazilian election campaigns.

Following a closed-door meeting, the court is anticipated to reaffirm the prohibition of deepfakes, whether they are used to support or attack a candidate. A key aspect of the upcoming ruling will be the establishment of clear criteria to differentiate deepfakes from other forms of AI-generated content, such as illustrations, caricatures, and animations, which may be permissible under certain TSE-imposed conditions. The primary factor for classification will be the degree of realism and verisimilitude, determining if the content is likely to mislead voters and create a false perception of authenticity.

Any precedent set by the TSE will be applied to all similar pending cases. Ministers are expected to debate whether the improper use of the former president's avatar warrants a penalty for Senator Flávio Bolsonaro, considering whether the convention itself constitutes an internal party act or a campaign activity. The ruling aims to provide clarity on the boundaries of AI use in political campaigning, balancing technological innovation with the integrity of the electoral process.