Overview
- The Electoral Prosecutor’s Office urged the TSE to limit AI use in campaign pieces to technical quality adjustments and to ban manipulations that create, substitute, omit, merge, accelerate or overlay images and audio, arguing disclosure labels are insufficient.
- Prosecutors proposed fines of R$5,000 to R$30,000 for disseminating AI‑fabricated or manipulated content that spreads blatantly false or gravely decontextualized claims, with penalties also possible for beneficiaries who had prior knowledge.
- Federal representatives asked to expand profile removals beyond bots or fake accounts, to avoid mass takedowns of informational government content, to prohibit chatbots from recommending candidates, and to place explicit risk‑mitigation duties on AI providers and open‑channel messaging apps.
- The data protection authority (ANPD) called for curbs on voter profiling and microtargeting in electoral advertising and for data‑protection impact assessments from providers that seek to boost paid political content.
- Debate over paid amplification intensified, as the government and PT pressed to bar boosting negative content in pre‑campaign periods and the Prosecutor’s Office sought to classify any paid promotion by individuals as premature propaganda.