ARTIFICIAL INTELLIGENCE AND DEMOCRACY:

LEGAL ANALYSIS OF THE USE OF DEEPFAKE IN THE ELECTORAL CONTEXT IN LIGHT OF TRE-MA JURISPRUDENCE

Abstract

The rapid advancement of artificial intelligence (AI) technologies has deeply transformed communication dynamics and electoral processes, with particular concern regarding the risks posed by deepfakes. While this audiovisual manipulation technique can be used for legitimate purposes, it also carries a high potential to undermine electoral integrity, social trust, and individual reputation. This article examines the intersection between artificial intelligence and democracy, focusing on Brazilian Electoral Court Resolution No. 23.732/2024 and a landmark decision of the Regional Electoral Court of Maranhão (TRE-MA), which addressed the dissemination of a satirical AI-generated video in Icatu-MA. The research combines normative and case law analysis, seeking to understand how Electoral Justice has balanced the prohibition of deepfakes with the protection of freedom of expression in political contexts. The study concludes that Resolution No. 23.732/2024 establishes a significant regulatory framework by imposing strict restrictions and mandatory transparency in the use of synthetic content, while case law reaffirms the need to distinguish harmful technological manipulations from legitimate political criticism, including satire. The findings highlight that the central challenge of contemporary electoral law lies in reconciling technological innovation, democratic integrity, and fundamental rights, requiring continuous regulatory updates and increasingly refined jurisprudential standards in the face of the growing sophistication of AI technologies.