Agentes Inteligentes e Curadoria da Pesquisa Acadêmica:

uma análise empírica de ferramentas de IA

Abstract

Abstract: This article seeks to empirically investigate the impact of generative artificial intelligence (GAI) on the curation of academic research. Based on a comparative study of the ChatGPT, DeepSeek, Consensus, and Research Rabbit tools, the research evaluates how these intelligent agents structure the search, organization, and validation of scientific information. The methodology adopted combines practical experimentation with qualitative analysis, exploring the transparency of sources, the accuracy of data, and the challenges of algorithmic mediation. The findings reveal both the potential of these technologies to optimize knowledge screening and their limitations, such as algorithmic bias, source opacity, and the risk of fictitious references. The results reinforce the need for media and algorithmic literacy as essential competencies for academic research in the AI ​​age.