Inteligência artificial e transformação da influência eleitoral em redes digitais

evidências de novas possibilidades na segmentação eleitoral na esfera pública digital brasileira para 2026

Abstract

This article examines how artificial intelligence (AI) is transforming the paradigm of digital influence in Brazilian electoral processes, with a prospective focus on the 2026 elections. The study is developed within the DR.IA laboratory, particularly through its Artificial Intelligence and Electoral Integrity Laboratory (AIEL–DR.IA), and aims to provide a comparative retrospective of the 2018 and 2022 elections. It investigates the transition from electoral microtargeting strategies—traditionally based on behavioral data analysis, to a new phenomenon termed generative influence. This concept refers to the capacity of AI systems not only to identify patterns within the electorate, but also to produce and deliver, at scale, at high speed (and virtually in real time), and without necessarily violating existing data protection standards, persuasive and highly personalized political content across multiple formats. The research is grounded in the premise that current data protection legislation is insufficient to mitigate the risks associated with this emerging architecture of AI-driven influence. It seeks to determine whether technological advances are reshaping the environment of digital political discourse, whether emerging forms of electoral segmentation can be identified, and whether new AI models can go beyond predicting and inducing political preferences based on discursive patterns. To address these questions, the study adopts an empirical approach, employing data science techniques such as topic modeling, sentiment analysis, and clustering on data collected from digital platforms (e.g., X and YouTube). The central hypothesis is that generative AI can reconfigure political communication by continuously optimizing, in real time and in response to behavioral signals, the production of narratives that maximize emotional and political engagement. An empirical understanding of this transformation is therefore essential to assess its impacts on electoral integrity and to inform the development of mechanisms aimed at strengthening and optimizing democratic citizenship.