CIEDDE and SIADE — Brazil's Electoral Court Coordinates a Multi-Agency Response to AI Deepfakes in the 2024 Municipal Elections

Brazil's Superior Electoral Court (TSE) built a layered response to AI-enabled electoral disinformation for the 2024 municipal elections.
The Integrated Center for Confronting Disinformation and Defending Democracy (CIEDDE), opened in 2024, brings the TSE together with six other public bodies, including the telecom regulator Anatel, the Federal Public Prosecutor's Office, the Ministry of Justice and the Federal Police, to centralise complaints and forward them to platforms. The Electoral Disinformation Alert System (SIADE) lets citizens report suspicious content; an automated system classifies it by topic and escalates indications of a crime to prosecutors and the Federal Police.
TSE Resolution 23.732 requires a disclaimer on any campaign content generated or edited by AI and bans synthetic audio and video that falsely depicts real or fictional persons, with penalties up to removal of a candidacy. In August 2024 the TSE signed memoranda with Meta, TikTok, LinkedIn, Kwai, X, Google and Telegram, with platforms to report on measures taken within two hours.
Evidence of effectiveness is thin. DFRLab and NetLab UFRJ monitoring (June-August 2024) found one deepfake through keyword searches, 16 media-reported cases across nine states and none in 1,588 WhatsApp or 854 Telegram groups, and warned that detection tools remain immature, Meta's ad library has no AI-content filter, and the platform agreements set no binding detection standards. The TSE's own commentators note uneven application across many municipalities and weaker effect where digital literacy is low. No takedown statistics were found in the sources reviewed.

Read the full analysis: https://international.tse.jus.br/en/misinformation-and-fake-news/elections-2024-ciedde