Artificial intelligence and co-authorship of scientific manuscripts: discussions about using ChatGPT in scientific research and writing

This paper intends to analyze the use of artificial intelligence tools, such as ChatGPT, in scientific research and writing and its conflicts with integrity standards in co-authorship. Thus, we aim to problematize the theme and suggest criteria for defining co-authorship in scientific works in the perspective of using artificial intelligence tools. It is concluded that, considering the current scenario, scientific journals should introduce policies on the issue, recommending that: 1. artificial intelligence mechanisms (such as ChatGPT) do not satisfy the requirements for co-authorship; 2. Authors must declare (in the methodology or in the acknowledgments) the use of artificial intelligence mechanisms in the writing of the article and/or in the other stages of the scientific research; 3. Authors are responsible for errors, plagiarism and other misconducts that eventually occur in their research due to the use of artificial intelligence mechanisms.