[REVISION HISTORY]

AI deepfake election threats in Latin America

Updated 2 times since CLSTR started tracking revisions of this situation.

What changed

2026-08-15 18:10 UTC → 2026-08-22 16:04 UTC · added ~~removed~~

In late July 2026, experts warned that generative AI was making deepfake content cheaper, faster and more convincing, posing new risks to electoral integrity across Latin America. Calls were made for transparent labeling, mandatory identification mechanisms and stronger verification tools, backed by a joint statement from over 60 privacy authorities urging developers to embed safeguards by design. A week later, electoral officials in the Dominican Republic and Panama outlined concrete steps to counter AI‑driven misinformation ahead of their 2028 and 2029 elections. The Dominican Republic’s communications chief, Danylsa Vargas, highlighted detection challenges, questioned whether the Central Electoral Board has sufficient protocols, and urged political parties to avoid manipulated media. In Panama, Electoral Tribunal director Yara Campos announced preparations to detect and sanction AI‑generated media. The tribunal is drafting rules requiring warning labels, a pre‑election ban on AI‑generated propaganda, and digital monitoring tools supported by the UN Development Programme, citing Brazil’s recent sanctions as a model. By mid-August, the focus expanded to Guyana, where experts noted an urgent need to modernize cyber laws to combat generative AI. Concerns have been raised that existing frameworks, such as the Cybercrime Act of 2018, are insufficient to address hyper-realistic fabricated audio and video. There are specific fears that synthesized content, including fake voice notes shared via WhatsApp, could exacerbate ethnic and political polarization within the country. In Brazil, judicial authorities have intensified warnings regarding technological threats. Supreme Federal Court Justice Cármen Lúcia rejected the notion of a ‘good deepfake’, arguing that AI-generated content is inherently false and capable of manipulating voter behavior even with subject authorization. Justice Alexandre de Moraes further warned that using AI to spread disinformation could lead to candidate disqualification, noting that existing regulations and the Penal Code will be used to hold responsible parties accountable.

Versions

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