AI Policy

Jump To Section:

Overview

We, the editorial team of the Journal of New Librarianship (JoNL), recognize that the use of generative artificial intelligence (AI) tools in the process of scholarly communication is an ongoing and rapidly changing issue that will require continuous review of journal policies and procedures. We understand that the evaluation, understanding, and teaching of AI tools is of utmost importance to the field of librarianship, and we encourage our authors to approach these topics with a sense of curiosity and a critical eye.

Accordingly, we request that our authors, peer reviewers, and editors follow the guidelines noted herein, with the understanding that this guidance may change as technology and our shared understanding of that technology evolves. In the face of a rapidly advancing technological environment, our intent is to ensure that research published in JoNL remains credible, transparent in its methodology, accountable in authorship, and reproducible to the greatest extent possible.

AI guidelines for use will likely continue to evolve as issues such as copyright, fair use, and research and AI ethics also evolve. As such, this policy will be reviewed yearly or as needed.

Should you have any concerns, questions, or issues regarding generative AI within JoNL publications, please contact newlibsjournal@gmail.com.

For Authors

General

At this time, we allow the use of generative AI as tools to support authors’ original research and submitted manuscripts. Manuscripts must be written by human authors and co-authors, and AI tools should only be used to support the author’s own ideation, critical thinking, and creative processes. Any use of generative AI must be disclosed as described in the Disclosure & Attribution section of this policy.

JoNL aligns with the COPE (Committee on Publication Ethics) position statement on authorship and ethical transparency in that authors must disclose what tool was used and how it was used. Additionally, AI cannot be credited as an author of a manuscript.

Authors are 100% responsible for the entirety of their manuscript, and must ensure that the inclusion of any content generated by AI tools is accurate, provides sufficient disclosure and attribution, and minimizes and acknowledges inherent biases.

Authors are prohibited from using AI tools to create falsified data and content, including text, data, images, etc., unless for the explicit purposes of illustrating a specific and stated argument.

U.S. law and guidance regarding copyright and generative AI is an evolving topic. Currently, the U.S. Copyright Office provides parameters for the use of generative AI content within copyrighted works through their policy, Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence. JoNL authors should check back frequently for updated information from the U.S. Copyright Office: Copyright and Artificial Intelligence.

In accordance with current U.S. Copyright guidance, AI-generated content is fair use and may be included within manuscripts. However, to be included within the copyright of a manuscript, authors must be the primary creators of the content and only use AI as a resource or tool, such as how one would use Adobe Photoshop to manipulate their existing image. Any use of generative AI must be disclosed within the manuscript as described in the Disclosure & Attribution section of this policy. As a reminder, authors should never use any copyrighted work in any capacity without permission or appropriate Creative Commons licensing. For further information regarding copyright, AI, and Creative Commons licensing, visit Understanding CC Licenses and Generative AI.

Data Privacy & AI Training Models

Authors should proceed with caution when feeding research data and/or authored content into generative AI tools, and always avoid inputting sensitive or confidential data. Data privacy and intellectual property ownership guidelines vary by tool and are murky at best; it’s possible that a tool creator may retain the right to use whatever content is fed into it.

While JoNL does not engage in or profit from the use of your work for AI training purposes, we acknowledge that openly accessible academic publications are frequently utilized by various entities for AI model training. This practice is widespread and often occurs without the explicit permission of authors or publishers. By publishing in our open access journal, you are contributing to the global dissemination of free knowledge. However, please be aware that this openness may also result in your work being incorporated into AI training datasets. We encourage authors to consider this information when deciding to publish.

Disclosure & Attribution

In furtherance of JoNL’s goals of credibility, transparency, accountability, and reproducibility, any and all use of AI tools within the process of research and writing must be disclosed, using APA 7 citation style, in two places:

  1. Within the body of the manuscript
  2. Within the References section of the manuscript

Depending on the context of use, authors may describe their use of AI within any area of the body of the manuscript, with the Methods section being a typical disclosure location.

Any chat transcript and/or text generated by AI should be submitted with your manuscript and follow the same formatting found within the manuscript template. Submit this material as follows:

In-Text Disclosure

  1. When disclosing AI use within the body of the manuscript, authors should specify*:
  2. The individual who used the generative AI tool
  3. The tool and version that was used
  4. The date of the use
  5. The prompt(s) used to generate the text
  6. The location in the manuscript readers can find the entire chat transcript (e.g., “For full chat transcript, see Appendix A.”)
  7. The sections(s) in the paper containing the content and/or ideas resulting from AI use.

*adapted from Hosseini, Rasmussen, & Resnik (2023)

Example Disclosure in Methods Section:

In writing this manuscript, [Author Initials] used Google Gemini on the 12th of February 2024. The following prompt was used to generate a number of complex Boolean search queries that were used in supporting the writing of the literature review section of this paper: “You are an academic librarian and expert researcher. You are using the Scopus database to find peer-reviewed scholarship on the concept of ‘One Health’ for the purposes of a literature review. Create a complex search string using Boolean operators, wildcards, and other tactics for finding a broad array of interdisciplinary research on the concept of One Health in the Scopus database. Similar concepts should be grouped together within the search and connect with the Boolean OR. Both technical terms or jargon as well as popular terms should be used in the search string.” The generated text was copied verbatim and is included as supplementary material.”

Example Disclosure in Body of Manuscript:

In response to the prompt “Help me ethically cite AI generated content that will be used within research,” the ChatGPT-generated text reminded me “...that ethical citation practices promote transparency and integrity in research, allowing others to assess and build upon your work accurately” (OpenAI, 2023). See Appendix B for full chat transcript.

Citation Formatting

APA 7 currently uses an adapted version of their reference template for software (APA, 2020, section 10.10) to cite AI generated content. For further information, visit APA Style Blog’s How to Cite ChatGPT.

Parenthetical In-Text Citation Formatting & Examples

(Author of the model, year)

(OpenAI, 2023)

(Google DeepMind, 2024)

Reference List Entry Formatting & Examples

Author of the model. (Version year of the tool). Title of the model (Model version) [Description of type of model]. Publisher. URL

Note: If the author and publisher are the same, omit the publisher name. This is the case for ChatGPT.

OpenAI. (2023). ChatGPT (Sept 25 version) [Large language model]. https://chat.openai.com/

Google DeepMind. (2024). Gemini (1.0 Pro version) [Large language model]. https://gemini.google.com/app/

Author Checklist

Prior to submitting a manuscript, authors should use this AI Policy Author Checklist to ensure they’ve met all of the criteria found in this policy.

For Peer Reviewers

In line with JoNL’s expectations of peer reviewers, peer reviewers are responsible for the integrity and accuracy of their submitted reviews, as well as responsible for upholding the utmost confidentiality of authors and their manuscripts. At this time, given the lack of standardized data privacy and intellectual property laws and guidelines regarding data submitted to AI tools, reviewers must refrain from inputting any portion of a manuscript into an AI tool (including AI detector tools), including their own recorded or otherwise reworded thoughts about the paper and/or inputting content for copyediting purposes. In addition, peer reviewers should refrain from using AI tools to conduct evaluations of a manuscript. All analyses should originate from and be shared by the peer reviewer themselves. Finally, given the unreliable and potentially biased nature of AI-detecting software, reviewers should refrain from using any sort of AI detectors with manuscripts.

For Editors

Editors are responsible for upholding the utmost confidentiality of authors and their manuscripts as well as peer reviewers and the feedback they provide. At this time, given the lack of standardized data privacy and intellectual property laws and guidelines regarding data submitted to AI tools, editors should refrain from inputting any portion of a manuscript into an AI tool (including AI detector tools), such as for the purposes of evaluation, summarization, or editorial assistance.

Acknowledgements

This policy has been adapted from the following scholarship: