The DOI matches its Crossref record; 6 of the 6 references it cites were found in a registry or an index; none of the 3 sources its claims cite is checked for support yet.
Toward Structured Transparency: A Governance Framework for AI Use in Biomedical Publishing
Open accessAbstract
The rapid adoption of large language models and generative artificial intelligence (AI) is transforming biomedical research and publishing.Although international organizations such as the International Committee of Medical Journal Editors (ICMJE) and the Committee on Publication Ethics (COPE) have established the principle that AI cannot be recognized as an author, and the ICMJE, in its January 2026 revision, has introduced a dedicated section addressing AI use by authors, peer reviewers, and editors, journal-level Instructions for Authors still require further operational detail to apply these principles consistently throughout the publication process.This review critically examines current AI-related policies in the Journal of Korean Medical Science (JKMS) and the Korean Association of Medical Journal Editors (KAMJE) and identifies three major challenges: the practical limitations of the AI non-authorship principle, the inadequacy of current AI disclosure practices, and the emergence of AI-specific conflicts of interest that extends beyond conventional financial disclosures.We argue that medical publishing should move from a restrictive approach toward a framework of structured transparency that systematically documents, evaluates, and verifies AI use.To achieve this goal, we propose a practical governance framework that includes a three-tiered AI disclosure system, strengthened accountability for corresponding authors, expanded institutional conflict-of-interest disclosures, transparent reporting of AI use by peer reviewers, and formal editorial policies governing AI-assisted editorial activities.Future revisions of the KAMJE and JKMS Instructions for Authors should prioritize transparent and accountable AI governance rather than restricting AI use.Adoption of a structured, publication-wide framework encompassing authors, peer reviewers, and editors would strengthen research integrity while supporting the responsible integration of AI into biomedical publishing.
Overview
CoChat Summary
TL;DR
This review argues that biomedical publishing should move from restrictive AI policies toward structured transparency and proposes a governance framework for authors, peer reviewers, and editors. It analyzes current JKMS and KAMJE AI-related policies and recommends a three-tiered AI disclosure system, strengthened corresponding-author accountability, expanded conflict-of-interest disclosure, reviewer AI disclosure, and editorial AI policies.
Key findings
- The paper identifies three major challenges: the practical limitations of the AI non-authorship principle, the inadequacy of current AI disclosure practices, and the emergence of AI-specific conflicts of interest that extend beyond conventional financial disclosures.
- It proposes a three-tiered AI disclosure framework: Tier 1 Assistive (language/grammar/translation/copy-editing/formatting) disclosed in Acknowledgments; Tier 2 Augmentative (AI-assisted drafting, literature synthesis, statistical programming, or analytical support) described in Methods with AI model and version, purpose of prompts, and access dates; Tier 3 Substantive (AI as substitute for human intellectual authorship) prohibited under routine practice, with exceptional cases requiring formal editorial review and independent verification.
- It recommends that corresponding authors certify all AI-assisted content via an Accountability Clause, that COI disclosure expand to institutional AI interests and collaborations/licensing/data-sharing with AI service providers, that peer reviewers disclose AI use under an analogous tiered framework with Tier 3 prohibited, and that editorial boards establish formal internal policies ensuring human oversight of editorial decision-making.
- It notes that the ICMJE’s January 2026 revision added a dedicated section on AI use by authors, peer reviewers, and editors, establishing that nondisclosure may in some circumstances be treated as misconduct and that journals should maintain a policy on AI use in review.
Methods
- Design: critical review and normative governance proposal, not an empirical study.
- Data or sample: current AI-related Instructions for Authors and policies of the Journal of Korean Medical Science (JKMS) and the Korean Association of Medical Journal Editors (KAMJE), with reference to ICMJE and COPE recommendations.
- Main analysis: identifies limitations in current guidelines and proposes recommendations for guideline reform, summarized in Table 1.
Limitations
- Stated limitations: Not stated in the paper.
- Evident from the text: the proposed framework is not empirically tested or validated, and no implementation data or outcomes are reported.
- Evident from the text: the analysis focuses on JKMS and KAMJE and the Korean publishing context, although the paper suggests the framework could serve as a practical reference for biomedical journals worldwide.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 Valid · 1 Not valid · 3 Not checked · 19 Own results and interpretations
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In these circumstances, institutional affiliations may introduce competing interests that extend beyond the personal financial interests of individual researchers.”
§ Toward Structured Transparency: A Governance Framework for AI Use in Biomedical Publishing > ANALYSIS OF CURRENT GUIDELINES AND EMERGING LIMITATIONS > Expanding the concept of conflicts of interest in the AI era
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The absence of operational criteria inevitably leads to inconsistent interpretations of acceptable AI use and authorship eligibility across journals.”
§ Toward Structured Transparency: A Governance Framework for AI Use in Biomedical Publishing > ANALYSIS OF CURRENT GUIDELINES AND EMERGING LIMITATIONS > Practical limitations of the AI non-authorship principle
boostedstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Rests on an assumption not shown here.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Moreover, without standardized reporting requirements, peer reviewers cannot adequately evaluate whether AI-generated content has been independently verified or whether potential hallucinations, algorithmic bias, or factual inaccuracies have been appropriately addressed.”
§ Toward Structured Transparency: A Governance Framework for AI Use in Biomedical Publishing > ANALYSIS OF CURRENT GUIDELINES AND EMERGING LIMITATIONS > Limitations of current AI disclosure practices
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
- Rests on an assumption not shown here.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Routine language editing or grammatical correction is treated equivalently to far more substantive uses, such as literature synthesis, statistical code generation, data interpretation, or scientific writing.”
§ Toward Structured Transparency: A Governance Framework for AI Use in Biomedical Publishing > ANALYSIS OF CURRENT GUIDELINES AND EMERGING LIMITATIONS > Limitations of current AI disclosure practices
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
- Rests on an assumption not shown here.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Without adequate documentation of AI involvement, reviewers and editors may be unable to determine whether questionable findings, unsupported interpretations, or factual inaccuracies originated from human judgment or AI-generated outputs.”
§ Toward Structured Transparency: A Governance Framework for AI Use in Biomedical Publishing > DIRECTIVES FOR GUIDELINE REFORM: TOWARD A FRAMEWORK OF STRUCTURED TRANSPARENCY
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: unclear — It is unclear whether it could be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
1 more claim
Create a free account to see all 23 claims and to verify the 2 references still open.
- It may state two separate claims as one. 9
- Probably not a substantial claim. 13
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Toward Structured Transparency: A Governance Framework for AI Use in Biomedical Publishing > RECOMMENDATIONS > Implement a three-tiered AI disclosure framework
“AI-assisted drafting of manuscript text should be reported in the Acknowledgments section, whereas AI used for data collection, analysis, or figure generation should be described in the Methods section.”
§ Toward Structured Transparency: A Governance Framework for AI Use in Biomedical Publishing > RECOMMENDATIONS > Establish institutional policies for editorial AI use
“| Authors | | Tier 2: Augmentative (structural and analytical assistance) | Describe in the Methods section, specifying the AI model & version, purpose of prompts, and access dates to enable review. |”
Citations
Cites
ICMJE (2024) · Ewha Medical Journal · doi:10.12771/emj.2024.e48 · Cited by 480
(2024) · doi:10.24318/ccvrzbms · Cited by 210
Defining the Boundaries of AI Use in Scientific Writing: A Comparative Review of Editorial Policies
Yoo, J.-H. (2025) · Journal of Korean Medical Science · doi:10.3346/jkms.2025.40.e187 · Cited by 91
Joo, H. J., Radnaabaatar, M., & Jung, J. (2026) · Journal of Korean Medical Science · doi:10.3346/jkms.2026.41.e306 · Cited by 7
Oh, J. (2026) · Journal of Korean Medical Science · doi:10.3346/jkms.2026.41.e298 · Cited by 5
Proper Utilization of Generative AI in Scholarly Writing: The Researcher and Author Perspective
Suh, D. H., & Han, D. S. (2026) · Journal of Korean Medical Science · doi:10.3346/jkms.2026.41.e304 · Cited by 4
Research works printed but not linked: 1
- International Committee of Medical Journal Editors (ICMJE). Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals. Updated 2026. Accessed 2026 July 14. https://www.icmje.org/recommendations/
Cited by
Comprehensive Consideration of Ethics in AI-assisted Scientific Writing and Peer Review
Yoo, J. (2026) · Journal of Korean Medical Science · doi:10.3346/jkms.2026.41.e281 · Cited by 3
Letter to the Editor: From AI Use to AI Governance: The Next Evolution of Peer Review
Ahn, K. S. (2026) · Journal of Korean Medical Science · doi:10.3346/jkms.2026.41.e346
Yoo, J. (2026) · Journal of Korean Medical Science · doi:10.3346/jkms.2026.41.e351
In the Age When AI Writes Papers, What Is Left for Researchers?
Yoo, J. (2026) · Journal of Korean Medical Science · doi:10.3346/jkms.2026.41.e352
Metadata
- Authors
- Kim, J.-M.
- Affiliations
- Seoul National University (KR)
- Corresponding author
- Kim, Jong-Min
- ORCID
- Kim, Jong-Min
- Venue
- Journal of Korean Medical Science
- Publisher
- XMLink
- Published
- 2026
- Volume
- 41
- Issue
- 29
- Pages
- e318-e318
- ISSN
- 1011-8934 (ISSN-L)
- 1598-6357
- PMC
- PMC13406945
- OpenAlex
- W7170179489
- Semantic Scholar
- 559c52d98eaa02e5e9b9b0f6af69a1bde4c75c01
- Cited by
- 4
- References
- 6
- Citations by year
- 2026: 4
- FWCI
- 17.80
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by-nc
- creativecommons.org/licenses/by-nc/4.0/ (tdm)
- Locations
- Journal of Korean Medical Science (publishedVersion, cc-by-nc)
- PubMed (publishedVersion)
- PubMed Central (publishedVersion, cc-by-nc)
- Topics
- Social Sciences › Library and Information Sciences › Publishing and Scholarly Communication
- Social Sciences › Education › Academic integrity and plagiarism
- Social Sciences › Library and Information Sciences › Academic Publishing and Open Access
- Keywords
- generative AI, large language models, AI disclosure, conflicts of interest, AI-assisted peer review, research integrity, journal editorial policy, ICMJE, CCOPE, AI governance framework
- MeSH terms
- Generative Artificial Intelligence, Artificial Intelligence, Authorship, Humans, Publishing, Conflict of Interest, Editorial Policies, Biomedical Research
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-11-12, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-11-12 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Toward Structured Transparency: A Governance Framework for AI Use in Biomedical Publishing. https://app.cochat.ai/papers/10.3346/jkms.2026.41.e318
- The paper
- Kim, J.-M. (2026). Toward Structured Transparency: A Governance Framework for AI Use in Biomedical Publishing. Journal of Korean Medical Science. https://doi.org/10.3346/jkms.2026.41.e318
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.
