Abstract
This policy has been adapted from the Sage Publishing: “ChatGPT and Generative AI” policy. 1
Sage and the Editors of the Journal of Children’s Orthopaedics (JCO) recognize the value of large language models (LLMs) (e.g. ChatGPT) and generative AI as productivity tools that can help authors in preparing their article for submission; to generate initial ideas for a structure, for example, or when summarizing, paraphrasing, language polishing, and so on. However, it is important to note that all language models have limitations and are unable to replicate human creative and critical thinking. Human intervention with these tools is essential to ensure that content presented is accurate and appropriate to the reader.
JCO therefore requires authors to be aware of the limitations of language models and to consider these in any use of LLMs in their submissions:
Objectivity: Previously published content that contains racist, sexist, or other biases can be present in LLM-generated text, and minority viewpoints may not be represented. Use of LLMs has the potential to perpetuate these biases because the information is decontextualized and harder to detect.
Accuracy: LLMs can “hallucinate,” that is, generate false content, especially when used outside of their domain or when dealing with complex or ambiguous topics. They can generate content that is linguistically but not scientifically plausible, they can get facts wrong, and they have been shown to generate citations that do not exist. Some LLMs are only trained on content published before a particular date and therefore present an incomplete picture.
Contextual understanding: LLMs cannot apply human understanding to the context of a piece of text, especially when dealing with idiomatic expressions, sarcasm, humor, or metaphorical language. This can lead to errors or misinterpretations in the generated content.
Training data: LLMs require a large amount of high-quality training data to achieve optimal performance. However, in some domains or languages, such data may not be readily available, limiting the usefulness of the model.
Guidance for authors
Authors are required to:
Clearly indicate the use of language models in the manuscript, including which model was used and for what purpose, in both the methods and acknowledgements sections.
Verify the accuracy, validity, and appropriateness of the content and any citations generated by language models and correct any errors or inconsistencies.
Provide a list of sources used to generate content and citations, including those generated by language models. Double-check citations to ensure they are accurate and are properly referenced.
Be conscious of the potential for plagiarism where the LLM may have reproduced substantial text from other sources. Check the original sources to be sure you are not plagiarizing someone else’s work.
Acknowledge the limitations of language models in the manuscript, including the potential for bias, errors, and gaps in knowledge.
Please note that AI bots such as ChatGPT should not be listed as an author on your submission.
We will take appropriate corrective action where we identify published articles with undisclosed use of such tools.
Authors should check the guidelines of the journal they are submitting to for any specific policies that may be in place on that journal.
General information on author responsibilities can be found on the Sage Author Gateway.
Guidance for editors and reviewers
Editors and reviewers should evaluate the appropriateness of the use of LLMs and ensure that the generated content is accurate and valid.
Further information
Please see the World Association of Medical Editors’ (WAME) recommendations on chat bots, ChatGPT, and scholarly manuscripts and the Committee on Publication Ethics’ (COPE) position statement on Authorship and AI tools.
This policy may evolve further as we work with our publishing partners to understand how emerging technologies can help or hinder the process of preparing research for publication. Please check back to this page for the latest information.