Loading...
Publication
Inteligência Artificial no auxílio à pesquisa científica: vantagens, desvantagens e uso responsável
Advisor(s)
Abstract(s)
A Inteligência Artificial (IA) vem ampliando de maneira significativa as práticas de pesquisa científica, ao apoiar processos de busca, leitura, síntese e escrita académica. Este estudo investigou como essas tecnologias podem fortalecer a produção científica sem comprometer a ética, a integridade e o protagonismo humano. A pesquisa seguiu abordagem qualitativa, aplicada e descritiva, fundamentada em revisão bibliográfica e documental, com análise de conteúdo. Os resultados revelam um ecossistema crescente de ferramentas, como Elicit, Perplexity, Scholarcy, Litmaps, Research Rabbit, NotebookLM e modelos generativos como o ChatGPT, que aceleram a revisão de literatura, melhoram a clareza textual, sintetizam múltiplas fontes e auxiliam em tarefas como tradução, elaboração de esboços e organização de ideias. Apesar das vantagens, observam-se riscos como plágio, inconsistências, dependência excessiva, fragilização de competências analíticas, variação das respostas a uma mesma solicitação e possíveis distorções metodológicas. Para mitigar tais riscos, são apresentadas diretrizes éticas que enfatizam autoria exclusivamente humana, transparência do uso da IA, validação cruzada das informações e a preservação da agência do pesquisador. Conclui-se que a IA deve ser utilizada como ferramenta complementar, capaz de ampliar a eficiência e a qualidade da produção científica, desde que seu uso seja crítico, responsável e alinhado a princípios éticos que mantenham o pesquisador como centro do processo de criação, interpretação e validação do conhecimento.
Artificial Intelligence (AI) has been significantly expanding scientific research practices by supporting processes such as searching, reading, synthesizing, and academic writing. This study investigated how these technologies can strengthen scientific production without compromising ethics, integrity, or human agency. The research adopted a qualitative, applied, and descriptive approach, grounded in bibliographic and documentary review, with content analysis. The results reveal a growing ecosystem of tools, such as Elicit, Perplexity, Scholarcy, Litmaps, Research Rabbit, NotebookLM, and generative models like ChatGPT, which accelerate literature reviews, enhance textual clarity, synthesize multiple sources, and assist with tasks such as translation, drafting, and idea organization. Despite these advantages, risks were identified, including plagiarism, inconsistencies, excessive dependence, weakening of analytical skills, variation in responses to the same query, and potential methodological distortions. To mitigate such risks, ethical guidelines are proposed, emphasizing exclusively human authorship, transparency in AI usage, cross-validation of information, and the preservation of researcher agency. It is concluded that AI should be employed as a complementary tool capable of enhancing the efficiency and quality of scientific production, provided its use is critical, responsible, and aligned with ethical principles that maintain the researcher at the center of knowledge creation, interpretation, and validation.
Artificial Intelligence (AI) has been significantly expanding scientific research practices by supporting processes such as searching, reading, synthesizing, and academic writing. This study investigated how these technologies can strengthen scientific production without compromising ethics, integrity, or human agency. The research adopted a qualitative, applied, and descriptive approach, grounded in bibliographic and documentary review, with content analysis. The results reveal a growing ecosystem of tools, such as Elicit, Perplexity, Scholarcy, Litmaps, Research Rabbit, NotebookLM, and generative models like ChatGPT, which accelerate literature reviews, enhance textual clarity, synthesize multiple sources, and assist with tasks such as translation, drafting, and idea organization. Despite these advantages, risks were identified, including plagiarism, inconsistencies, excessive dependence, weakening of analytical skills, variation in responses to the same query, and potential methodological distortions. To mitigate such risks, ethical guidelines are proposed, emphasizing exclusively human authorship, transparency in AI usage, cross-validation of information, and the preservation of researcher agency. It is concluded that AI should be employed as a complementary tool capable of enhancing the efficiency and quality of scientific production, provided its use is critical, responsible, and aligned with ethical principles that maintain the researcher at the center of knowledge creation, interpretation, and validation.
Description
Keywords
ChatGPT Ética na pesquisa Inteligência artificial na pesquisa Pesquisa científica
Pedagogical Context
Citation
Ramos, H.; Dala, M. A. dos S.; Gonçalves, B. M. F. (2025). Inteligência artificial no auxílio à pesquisa científica: vantagens, desvantagens e uso responsável. Revista Interdisciplinar de Pesquisas Aplicadas. ISSN 2675-6552. 1:2, p. 81–104
Publisher
Sociedade Espírito-Santense de Pesquisas de Tecnologias em Desenvolvimento