Background and Purpose
The rapid development of generative artificial intelligence (Generative AI) technologies, such as large language models (LLMs), has created new opportunities as well as new challenges in the process of writing, editing, and reviewing scholarly manuscripts. SHEM considers it necessary to establish a clear policy on the use of Generative AI by authors, editors, and reviewers in order to safeguard scientific integrity, transparency, and accountability throughout the manuscript handling and publication process.
This policy is developed with reference to the principles established by the Committee on Publication Ethics (COPE) and common practices adopted by international scholarly publishers.
Scope and Definitions
Generative AI, for the purposes of this policy, refers to AI-based tools capable of producing text, images, code, data, or other content in response to user prompts, including but not limited to ChatGPT, Claude, Gemini, Copilot, and similar tools.
This policy applies to all stages of the SHEM editorial workflow, including manuscript preparation, peer review, editing, and final production of the article.
Provisions for Authors
Authors are permitted to use Generative AI to:
- Improve grammar, spelling, and sentence clarity (language polishing).
- Assist in summarizing literature as a starting point for exploration, subject to manual verification against original sources.
- Assist with programming code or data analysis, provided the results are verified by the authors.
- Assist in drafting an initial outline of the manuscript.
Authors are prohibited from:
- Listing Generative AI as an author or co-author, as AI cannot bear scientific responsibility and accountability as required of authors.
- Using Generative AI to generate research data, experimental results, or findings that were not genuinely obtained through a legitimate research process.
- Using Generative AI to generate citations or references without verifying that they exist and are accurate.
- Submitting substantial parts of the manuscript (analysis, discussion, or conclusions) generated entirely by AI without the authors' intellectual contribution and verification.
If Generative AI is used in the writing process, authors must include a disclosure statement, in the Acknowledgment or Methods section, covering:
- The name and version of the AI tool used.
- The part(s) of the manuscript the AI assisted with (e.g., language editing, data analysis).
- Confirmation that the authors have reviewed, edited, and take full responsibility for the entire content of the manuscript, including AI-assisted sections.
Full Responsibility: Full responsibility for the originality, accuracy, and integrity of the manuscript remains with the authors, regardless of the extent to which Generative AI was used in its preparation.
Provisions for Reviewers
Reviewers are prohibited from uploading the manuscript, or any part of it, to public Generative AI platforms (including for the purpose of summarizing or translating). Doing so may breach manuscript confidentiality and the authors' intellectual property rights, as data uploaded to most public AI platforms may be stored or used by third parties for model training.
Reviewers may use Generative AI in a limited capacity to:
- Improve the grammar of their own review comments, without including excerpts from the manuscript under review.
- Understand general technical terminology that is not directly related to the confidential content of the manuscript.
Scientific evaluation, methodological assessment, and editorial recommendations must be the reviewer's own judgment. Generative AI must not be used to generate the substance of an assessment or recommendation regarding a manuscript.
Provisions for Editors
Editors are prohibited from uploading authors' manuscripts to public Generative AI platforms in any form during the editorial process, including for summarization, translation, or preliminary screening.
Editors may use Generative AI to:
- Draft routine correspondence with authors or reviewers (invitations, reminders, administrative communication) that does not contain substantive manuscript content.
- Assist with language editing of general journal pages (policies, author guidelines, informational pages).
Decisions to accept, request revision of, or reject a manuscript must be based entirely on the editor's own judgment, informed by reviewer assessments and editorial evaluation — not on recommendations generated by Generative AI.
Editors may use AI-content detection tools as an auxiliary screening aid; however, detection results alone must not be used as the sole basis for rejecting a manuscript without further investigation, given the current limited accuracy of AI detection tools.
Violations and Consequences
Non-compliance with this policy — including but not limited to failure to disclose AI use, listing AI as an author, or uploading confidential manuscripts to public AI platforms — will be addressed in accordance with SHEM's publication ethics guidelines and may result in manuscript rejection, a request for revision, or other actions consistent with COPE guidelines.
Policy Review
This policy will be reviewed periodically in line with the evolving capabilities of Generative AI technology and best practices adopted by the international scholarly publishing community.
This policy takes effect as of its date of publication and forms an integral part of SHEM's Publication Ethics guidelines.
References and Guidelines
This policy has been formulated in accordance with established international ethical frameworks and guidance regarding AI usage in academic publishing, including:
- International Association of Scientific, Technical and Medical Publishers (STM): Recommendations for a Classification of AI Use in Academic Manuscript Preparation.
- Elsevier: Policy on the Use of Generative AI and AI-Assisted Technologies in Writing and Peer Review.
- World Association of Medical Editors (WAME): Recommendations on Chatbots, Generative AI, and Scholarly Manuscripts.
