Use of AI Tools in Scholarly Publications

General Publication Guidance

Updated: 10/25/2023

Per HMS guidance: Consult journals regarding AI tool usage in research publications but, in general, follow the standards from the Committee on Publication Ethics (COPE) and the World Association of Medical Editors (WAME).

The field of AI is continually evolving, and users are encouraged to seek out the most current resources and guidelines to ensure compliance with all publication standards. The following is a summary of COPE and WAME guidance, sourced 10/25/2023.

• Only Humans can be Authors
  • Rationale: Chatbots do not meet the International Committee of Medical Journal Editors (ICMJE) authorship criteria. They are not legal persons, cannot approve the published version, and are not accountable for the accuracy and integrity of the work.
  • Legal Standing: AI tools cannot understand a conflict-of-interest statement or have the legal standing to sign.

• Authors Should Acknowledge the Sources of their Materials
  • If Chatbot/AI is Used to Draft:
    • General guidance is to note in the acknowledgment. 
    • Specify the prompts used to a) generate text or b) convert text or text prompts into tables or illustrations.
      • Some publications impose strict regulations on the use of AI-generated images. Consult the publication for their specific policies; this advisory is broadly applicable to all guidance provided here.
  • If Used for Analytical Work, Reporting Results, or Writing Computer Codes:
    • State in the body of the paper (Abstract and Methods).
    • Provide the full prompt to generate research results, the time and date of the query, and the AI tool with the version utilized.

• Authors Must Take Public Responsibility for Accuracy and Proper Attribution
  • As standard for any resource, be sure to validate the content and appropriately attribute all sources for AI-generated material. 
  • AI-generated content can be inaccurate, misleading, entirely fabricated (sometimes called “hallucinations”), or may contain copyrighted material. Review your AI-generated content before publication.

• Consult with the Publisher for their Editorial Policy
  • Guidance varies between publications, particularly on the use of generative images.
  • Link to Editorial Policies:
    • Nature Portfolio
    • New England Journal of Medicine
    • Science Journals
    • The Lancet
    • JAMA Network
    • Elsevier (Cell Magazine)

References

Zielinski C, Winker MA, Aggarwal R, Ferris LE, Heinemann M, Lapeña JF, Pai SA, Ing E, Citrome L, Alam M, Voight M, Habibzadeh F, for the WAME Board. Chatbots, Generative AI, and Scholarly Manuscripts: WAME Recommendations on Chatbots and Generative Artificial Intelligence in Relation to Scholarly Publications. WAME, World Association of Medical Editors. May 31, 2023. https://wame.org/page3.php?id=106

Authorship and AI tools: COPE Position Statements. COPE, Committee on Publication Ethics. February 13, 2023, https://publicationethics.org/cope-position-statements/ai-author

The information provided on this page is intended to offer an overview of general publication guidance. Please note that the field of artificial intelligence is continually evolving, and the capabilities of AI tools are rapidly advancing. As such, the information provided here may become outdated, and users are encouraged to seek out the most current resources and guidelines.

Users are responsible for ensuring that their use of AI tools in scholarly publishing adheres to all ethical guidelines, standards, and policies. The responsibility for the accuracy, integrity, and proper attribution of the work lies with the authors, and users should diligently verify the output of AI tools and ensure proper acknowledgment and attribution of AI-generated content.

This page does not endorse any specific AI tool or chatbot, and the inclusion of any tool or example does not imply endorsement. Users should exercise their own judgment and consult with publishers and ethical guidelines to ensure compliance with all publication standards.