Artificial Intelligence and Writing Assisted by Algorithms: What Do Scientific Associations and Their Journals Signal?

Abstract

The use of artificial intelligence (AI), particularly in the writing of scientific texts, has been much problematized. This article analyzes, in research of an exploratory and documentary nature (Lüdke; André, 1988), whether there are positions taken by Brazilian scientific associations and their respective journals on the production of scientific communications mediated by the use of artificial Intelligence (AI), such as ChatGPT and like, unveiling the emerging themes associated with the practices of appropriation of this technology in textual production. The research was carried out between May and June 2023, on the websites of 33 Brazilian scientific associations from different areas of knowledge, and 50 journals. No expressly contrary positions were found in relation to the uses of generative AI for textual production, by the 33 (thirty-three) associations researched. Only 3 (three) of the 50 (fifty) journals, until June 2023, had explicit policies or guidelines for authors on the use of artificial intelligence in the writing of scientific communication. From the analytical induction, when exploring the collected data, we conclude that: a) topics such as authorship, plagiarism and ethics are the most correlated with the appropriation of AI for scientific communication; b) the absence of positions from a significant number of journals, on the use of generative AI in the textual production of scientific communication is still safeguarded by the principles of authorship and ethics, with no general mention of the technologies that generate writing; c) concerns about the use of AI in scientific communication tend to spread, since its effects are not restricted to a specific type of knowledge. It is necessary for scientific associations and journals to explain their positions on the use, or not, of AI in scientific communication.