Policies on the use of Artificial Intelligence (AI) tools

The following policies for the use of AI tools are based on recommendations from the Committee on Publication Ethics (COPE) and best practices in scientific-academic publishing.

1. AI tools cannot be considered authors or co-authors because they cannot assume responsibility for the content produced, cannot make ethical decisions, cannot indicate the presence or absence of conflicts of interest, and cannot manage copyrights or license agreements.

2. Author responsibility

Authors are responsible for the originality of all content generated or assisted by AI tools.

Authors must review and verify content generated using AI tools to avoid errors, biases, or inaccuracies; they must also ensure that such content does not constitute plagiarism, fabrication, or falsification of data, nor infringe copyrights.

Additionally, the author must verify the accuracy of citations and references included in content generated by AI tools.

Articles in which AI tools have previously been applied to review spelling or writing are not allowed, because when reviewing the originality of the work, most or all of its content would be identified as generated by such tools. Additionally, the content of the manuscript and instructions for its processing in these AI tools may become training data for these tools, which are retained and may be used to generate new responses in the future, creating a risk of unintentional plagiarism by people using AI tools, as well as affecting the confidentiality of information and the article review process.

3. Declaration of the use of AI tools

Authors must explicitly declare the use of AI tools in the development of their work:

  • Their use must be indicated in the Methodology section of the article; in other types of articles, this will be indicated in the Introduction.
  • Authors must report whether they used AI tools for:
  • Collection, systematization, synthesis, and analysis of information
  • Generation of results
  • Contrasting differences or similarities between the theoretical information and the results obtained
  • Explaining the results obtained
  • Complementing the development of the discussion section
  • Preparation of tables, figures, images, audio, videos, etc.
  • Authors must also indicate which AI tools were used for the above, as well as the instructions (prompts) used to generate the results obtained.
  • The citation and references for the AI tools used in the article will be made in accordance with the American Psychological Association (APA, 2023) format, as described below:

a. Example of citation*

When asked: “Is the left-brain/right-brain divide real or a metaphor?”, the text generated by ChatGPT indicated that although the two cerebral hemispheres are somewhat specialized, “the notion that people can be characterized as ‘left-brained’ or ‘right-brained’ is considered an oversimplification and a popular myth” (OpenAI, 2023).

In the example above, the instruction (prompt) corresponds to the question posed to ChatGPT.

2. In-text citation format*

  • Parenthetical citation: (OpenAI, 2023)
  • Narrative citation: OpenAI (2023)…

c. Reference model for AI tools according to APA format*

OpenAI. (2023). ChatGPT (Mar 14 version) [Large language model].

https://chat.openai.com/chat (https://chat.openai.com/chat)

The structure of the reference for an AI tool according to APA format is as follows:

  • Author: the author of the model is OpenAI
  • Year: include only the year in parentheses: (2023)
  • Title: the name of the model in italics: ChatGPT
  • Version number: in parentheses and including its date: (Mar 14 version)

            o Other large language models or software may indicate the version differently; in that case, use the                    version format employed by the model’s author or publisher

  • Model type: the descriptor of the model in brackets: [Large language model]
  • Source:

                 o In this case, if the publisher’s name and the author are the same (as with ChatGPT), indicate the                   URL directly:

https://chat.openai.com/chat (https://chat.openai.com/chat)

  • The URL should lead as directly as possible to the page where the model is accessed, not to the model author’s homepage.
  • If the names of the author and publisher are different, first indicate the publisher, then the URL leading to the model.

*NOTE: this example is a template for citing and preparing references, according to APA format, for AI tools such as DALL·E 3, Perplexity AI, CANVA AI, among others.

Source:

McAdoo, T. (2023, April 7). How to cite ChatGPT. APA Style.

https://apastyle.apa.org/blog/how-to-cite-chatgpt

4. Appropriate and ethical use of AI tools

The use of AI tools must not compromise ethics, avoiding data manipulation, plagiarism, or infringement of intellectual property.

5. Declaration of the use of AI tools by editorial teams and reviewers

The editorial team may use AI tools for editorial (e.g., plagiarism detection) and administrative tasks, which can be carried out under strict human supervision.

Reviewers may not use AI tools. The use of AI tools in manuscript review compromises the confidentiality of the process and affects authorship rights: the information in the articles, as well as the instructions for processing them, may become training data for such tools, which are retained and may be used to generate new responses in the future, creating a risk of unintentional plagiarism by people who use AI tools.

6. Consequences for non-compliance with editorial policies on the use of AI tools

Non-compliance with editorial policies on the use of AI tools leads to sanctions such as the withdrawal of the submitted work.

7. Updating editorial policies on the use of AI tools

Editorial policies will be reviewed periodically to update and adapt them to technological changes in AI tools and best practices in scientific publishing, helping to maintain their relevance and pertinence.