Artificial Intelligence (AI) Policy

This policy applies to all authors, reviewers, editors, editorial staff, and other contributors involved in the submission, peer review, editing, production, and publication of manuscripts in COSIE.

The purpose of this policy is to promote responsible and transparent use of AI-assisted technologies while protecting research integrity, confidentiality, originality, intellectual responsibility, and the reliability of scholarly communication.

Authors must disclose substantive use of AI-assisted technologies in manuscript preparation, research, analysis, or production.

Disclosure is required when AI is used for:

• generating or substantially rewriting manuscript text;
• developing substantive research content;
• data analysis or interpretation;
• generating or modifying research code;
• generating tables, figures, diagrams, or graphical abstracts;
• generating synthetic data;
• substantive translation;
• literature analysis or research synthesis; and
• other substantive activities that influence the research or manuscript.

Authors should identify the AI tool used, its purpose, and the nature and extent of its use.

Artificial Intelligence systems, including ChatGPT, Gemini, Claude, Copilot, or other AI tools, cannot be listed as authors or co-authors of a COSIE manuscript.

AI systems cannot assume responsibility for the accuracy, integrity, originality, ethical compliance, or final approval of scholarly work. Authorship and accountability therefore remain with human researchers.

Authors remain fully responsible for all content submitted to COSIE, including material produced or assisted by AI.

Authors must critically review and verify AI-generated outputs, because AI systems may produce inaccurate, incomplete, biased, misleading, or fabricated information.

• Research findings
• Statistical and mathematical results
• Computer code
• References and citations
• Tables and figures
• Factual claims
• AI-generated text and images

Authors must not rely on AI-generated references without verification. AI systems may generate incorrect, incomplete, or nonexistent bibliographic information.

Every reference included in a manuscript must correspond to a real and verifiable scholarly source. Authors must verify the original source and ensure that the citation accurately supports the corresponding statement.

AI-generated material must not be used as a substitute for appropriate scholarly sources.

The use of AI does not exempt authors from COSIE's policies on plagiarism, duplicate publication, fabrication, falsification, copyright infringement, or other forms of research misconduct.

• AI must not be used to fabricate research data.
• AI must not be used to fabricate experimental results.
• AI must not be used to create fictitious references.
• AI must not be used to manipulate research evidence misleadingly.
• Authors remain responsible for plagiarism and proper attribution.

Authors, reviewers, and editors must protect confidential information. Submitted manuscripts and other privileged editorial materials must not be uploaded to external AI systems where confidentiality cannot be assured.

Examples include:

• unpublished manuscripts;
• confidential research data;
• personal or identifiable participant information;
• patient or health-related information;
• proprietary company information;
• confidential source code;
• unpublished research results; and
• confidential third-party documents.

When AI or machine learning constitutes an actual component of the research methodology, its use must be described as part of the research method rather than treated merely as editorial assistance.

Where applicable, authors should report:

• AI model or system;
• tool and version;
• purpose of use;
• input data;
• important parameters;
• preprocessing procedures;
• evaluation procedures; and
• other information necessary for reproducibility.

The use of generative AI to create or substantially modify images, illustrations, diagrams, graphical abstracts, or other visual materials must be disclosed where applicable.

AI must not be used to fabricate or manipulate research images, experimental evidence, or other research outputs in a way that misrepresents the underlying research.

Reviewers must maintain the confidentiality of manuscripts and editorial materials received for peer review.

Reviewers must not upload submitted manuscripts, confidential reviewer reports, unpublished research findings, or other confidential editorial information into public AI systems where confidentiality cannot be assured, unless such use is explicitly permitted by the journal.

If reviewers use AI in the review process where permitted, they should disclose such use to the journal and remain responsible for the accuracy and appropriateness of the review.

Editors must protect the confidentiality of submitted manuscripts and editorial correspondence.

Editors should not upload manuscripts or confidential editorial materials into external AI systems where confidentiality cannot be guaranteed.

Editorial decisions, including decisions concerning acceptance, revision, rejection, and peer review, remain the responsibility of human editors.

Authors who have made substantive use of AI-assisted technologies should include an AI declaration in the manuscript, preferably before the References section.

The declaration should identify the AI tool, its purpose, and the nature of its use.

During the preparation of this manuscript, the authors used [NAME OF AI TOOL] to assist with [describe the purpose]. The authors reviewed and verified all AI-assisted outputs and take full responsibility for the accuracy, originality, integrity, and final content of the manuscript.

The authors declare that no generative artificial intelligence or AI-assisted technologies were used in the preparation, analysis, or writing of this manuscript.

Failure to disclose substantive AI use may constitute a breach of publication ethics where disclosure is required by this policy.

If undisclosed AI use or inappropriate AI-assisted content is identified, COSIE may investigate the matter and take appropriate editorial action in accordance with applicable publication ethics procedures.

Depending on the circumstances, editorial action may include a request for clarification or correction, rejection of a manuscript, an editorial notice, an expression of concern, or appropriate post-publication action.

1. International Committee of Medical Journal Editors (ICMJE). Use of Artificial Intelligence in Publishing.
   https://www.icmje.org/recommendations/browse/artificial-intelligence/
2. International Committee of Medical Journal Editors (ICMJE). Use of AI by Authors.
   https://www.icmje.org/recommendations/browse/artificial-intelligence/ai-use-by-authors.html
3. International Committee of Medical Journal Editors (ICMJE). Use of AI by Reviewers.
   https://www.icmje.org/recommendations/browse/artificial-intelligence/ai-use-by-reviewers.html
4. International Committee of Medical Journal Editors (ICMJE). Editors' Role in Ensuring Responsible Use of AI.
   https://www.icmje.org/recommendations/browse/artificial-intelligence/ai-use-by-editors.html
5. World Association of Medical Editors (WAME). Chatbots, Generative AI, and Scholarly Manuscripts: WAME Recommendations on Chatbots and Generative Artificial Intelligence in Relation to Scholarly Publications.
   https://www.wame.org/page3.php?id=106

This policy is based on internationally recognized principles concerning transparency, human accountability, research integrity, confidentiality, and responsible use of AI-assisted technologies. COSIE may update this policy as AI technologies and international publishing standards evolve.