KMF PUBLISHERS
POLICY AND GUIDELINES ON RESPONSIBLE USE OF ARTIFICIAL INTELLIGENCE IN SCHOLARLY PUBLISHING

Responsible AI • Human Scholarship • Transparency • Research Integrity • Accountability

Effective Date: September 2026
Policy Version: 1.0
Policy Status: Publisher-wide Policy
Applies to: Journals, Books, Book Chapters, Conference Publications and Other Scholarly Content Published by KMF Publishers

1. Preamble

Artificial Intelligence (AI), particularly generative artificial intelligence (GenAI), large language models (LLMs), multimodal AI systems, AI research assistants, AI agents, deep-research systems, machine-learning applications, and AI-assisted analytical and creative technologies, is rapidly transforming research and scholarly communication.

KMF Publishers recognizes that responsible use of AI can enhance research efficiency, accessibility, language quality, data analysis, visualization, discovery, and scholarly communication. At the same time, inappropriate or undisclosed use of AI may create serious risks relating to research integrity, fabricated information, false citations, plagiarism, copyright, confidentiality, privacy, bias, authorship, reproducibility, and accountability.

KMF Publishers therefore adopts a principle of Responsible AI with Human Intellectual Leadership.

AI may assist scholarship. It must not replace scholarly responsibility.

The central principle of this policy is: “AI may assist the scholarly process, but human authors must remain the intellectual originators, critical evaluators, decision-makers, and accountable owners of the published work.”

2. Purpose

This policy establishes a transparent and proportionate framework governing the use of AI throughout the scholarly publishing lifecycle. It is intended to protect research integrity while allowing researchers to benefit responsibly from emerging technologies.

The policy applies not only to manuscript writing but also, where relevant, to research design, literature discovery, data processing, coding, statistical analysis, visualization, translation, image generation, peer review, editorial assessment, and post-publication scholarly communication.

3. Scope

This policy applies to all authors, co-authors, editors, Editors-in-Chief, editorial board members, guest editors, peer reviewers, book editors, chapter contributors, translators, research collaborators, production personnel, and other persons participating in publications under KMF Publishers.

It covers generative AI systems, large language models, AI writing assistants, AI research assistants, AI agents, deep-research tools, machine-learning systems, generative image and multimedia systems, AI coding assistants, AI translation systems, and comparable technologies capable of generating, transforming, interpreting, analysing, or recommending scholarly content.

The policy is technology-neutral. It therefore applies to current and future AI systems irrespective of brand, platform, model, or provider.

4. The KMF Human Intellectual Leadership Principle

KMF Publishers does not prohibit responsible AI assistance. However, every publication must demonstrate meaningful human intellectual leadership.

Human authors must determine the research problem, exercise scholarly judgment, critically evaluate evidence, verify factual claims, assess the reliability of sources, interpret findings, approve conclusions, and accept responsibility for the final publication.

A manuscript must not become merely an uncritical reproduction of AI-generated material. The greater the intellectual role played by AI, the greater the obligation for transparency, documentation, verification, and human oversight.

5. AI Cannot Be an Author

AI systems, chatbots, language models, algorithms, software applications, or AI agents cannot be listed as authors or co-authors of publications issued by KMF Publishers.

Authorship requires capacities that AI systems cannot independently exercise, including responsibility, consent, accountability, management of conflicts of interest, approval of the final manuscript, and response to questions concerning research integrity.

All listed authors therefore remain fully responsible for the publication regardless of whether AI contributed to any part of the research or manuscript-preparation process.

6. KMF Three-Level AI Use Classification

To make AI disclosure understandable and proportionate, KMF Publishers adopts a three-level classification. The classification is based on the function performed by AI, rather than merely the name of the software used.

Classification	Typical Uses	KMF Requirement
Level I – AI-Assisted	Spelling, grammar, punctuation, formatting, minor language refinement, reference-style formatting	Normally permitted without formal disclosure.
Level II – AI-Supported Scholarly Work	Substantive rewriting, translation, idea exploration, literature discovery, coding assistance, data analysis, visualization, explanatory figures, summarization, structural assistance	Permitted with appropriate human verification and disclosure.
Level III – AI-Integrated Research	AI forming a material part of research design, data collection, modelling, classification, prediction, qualitative coding, systematic-review workflow, algorithmic analysis, experimental methodology or interpretation	Permitted where methodologically and ethically appropriate; detailed methodological disclosure and reproducibility information are required.

7. Permitted AI-Assisted Uses

Authors may use AI responsibly to support language improvement, brainstorming, organization, translation, literature discovery, coding, statistical or computational assistance, data visualization, development of explanatory diagrams, preparation of plain-language summaries, and other legitimate scholarly activities.

Such use is permissible only when appropriate human oversight is maintained. AI must enhance—not substitute for—the author’s scholarly expertise and critical reasoning.

8. Uses Requiring Disclosure

Substantive AI involvement must be disclosed.

Disclosure is generally required where AI materially contributes to text generation or rewriting; translation; literature identification or synthesis; research design; coding; statistical analysis; qualitative analysis; data classification; visualization; creation of explanatory images or graphical abstracts; systematic-review processes; or other intellectual or methodological components of the research.

Where AI forms part of the research methodology itself, its use must additionally be described in the Methods/Methodology section with sufficient detail to permit scholarly evaluation.

9. Uses Normally Not Requiring Disclosure

Routine technologies used solely for spelling correction, punctuation, basic grammar checking, conventional reference formatting, or minor typographical improvements normally do not require an AI declaration, provided they do not materially generate or transform scholarly content.

If a tool substantially rewrites, reorganizes, interprets, generates, or changes the intellectual meaning of content, the use should be disclosed. When uncertain, KMF Publishers encourages disclosure.

Transparency will not in itself disadvantage a manuscript.

10. Mandatory Human Verification

Authors must independently review and verify all substantive AI-assisted or AI-generated outputs before submission.

Human verification must include factual accuracy, logical consistency, methodological validity, calculations, statistical interpretation, quotations, citations, references, data interpretation, originality, copyright compliance, and possible bias.

The use of AI does not transfer responsibility from the authors to the AI provider or to KMF Publishers. The corresponding author assumes responsibility for ensuring that all authors are aware of and agree with the declared use of AI.

11. Zero Tolerance for Fabricated References

KMF Publishers considers fabricated or unverifiable references a serious research-integrity concern.

AI-generated citations must never be accepted solely because they appear plausible. Authors must independently verify every reference used in a manuscript against the original scholarly source wherever reasonably possible. DOI, author names, article or book title, journal or publisher, year, volume, issue, and page/article number should be checked as applicable.

Submitting systematically fabricated references, false DOIs, invented quotations, or nonexistent sources may result in rejection, correction, investigation, or retraction depending on the circumstances.

12. AI and Literature Reviews

AI may support literature discovery, organization, screening assistance, keyword development, or preliminary synthesis. However, authors remain responsible for determining the scholarly relevance and reliability of sources and for reading and accurately representing the literature on which substantive claims depend.

For systematic reviews, scoping reviews, meta-analyses, bibliometric studies, and other evidence-synthesis research, any material AI involvement in searching, screening, extraction, coding, analysis, synthesis, or interpretation must be transparently described in the methodology.

AI must not silently replace a reproducible scholarly search or evidence-selection process.

13. AI in Research Methodology

Where AI is an integral research instrument, its use is legitimate when scientifically justified and ethically appropriate.

Authors should report, where relevant, the AI tool or model, version or identifiable release, provider, purpose, relevant settings or parameters, date or period of use, nature of human oversight, validation procedures, and sufficient workflow information to permit evaluation or replication.

Where prompts or AI workflows materially affect reproducibility, authors should preserve appropriate records and may be requested to provide them confidentially to the editorial office.

KMF Publishers does not automatically require publication of every prompt. The level of disclosure should be proportionate to its importance for understanding or reproducing the research.

14. AI, Data and Statistical Analysis

AI may assist data analysis when its use is scientifically appropriate. Authors must nevertheless retain responsibility for the validity of the dataset, analytical choices, assumptions, statistical procedures, interpretation, and conclusions.

AI must never be used to invent observations, participants, experimental results, survey responses, interview responses, numerical results, statistical significance, or other research evidence.

Synthetic data may be used where scientifically justified, but it must be clearly identified as synthetic and its generation methodology must be reported.

15. AI-Generated Images, Figures and Visual Content

AI-generated explanatory illustrations, conceptual diagrams, graphical abstracts, or similar non-evidentiary visual materials may be acceptable when relevant, lawful, and transparently disclosed.

Authors must identify AI-generated visual material where its origin would otherwise be unclear and should state the tool used when appropriate.

AI must not be used to fabricate or deceptively manipulate primary research images or evidentiary materials, including clinical images, microscopy images, diagnostic images, experimental photographs, documentary evidence, or other visual research data in a manner that misrepresents the underlying evidence.

Routine, scientifically justified image processing must follow relevant disciplinary standards.

16. AI and Coding

AI-assisted code generation or debugging is permitted. Authors remain responsible for verifying the accuracy, security, licensing implications, reproducibility, and methodological validity of AI-assisted code.

Where AI-generated or AI-modified code materially contributes to the research method or results, this should be disclosed in the methodology or supplementary information as appropriate.

17. AI Translation

AI-assisted translation is permitted when followed by competent human verification. Authors remain responsible for ensuring that the translated version accurately represents the meaning, terminology, evidence, and conclusions of the original material.

Substantive AI translation should normally be disclosed.

18. Privacy, Confidentiality and Sensitive Data

Authors must not upload confidential, proprietary, personally identifiable, sensitive, restricted, copyrighted, unpublished third-party, or ethically protected information to external AI systems unless they have lawful authority and appropriate safeguards to do so.

Particular caution is required for patient information, participant data, interview transcripts, confidential institutional records, commercially sensitive information, and unpublished research belonging to others.

Authors are responsible for understanding the relevant AI provider’s privacy, data-retention, intellectual-property, and model-training terms before uploading protected material.

19. Copyright and Intellectual Property

Use of AI does not exempt authors from copyright, licensing, attribution, plagiarism, or intellectual-property obligations.

Authors must ensure that AI-assisted or AI-generated material does not unlawfully reproduce protected material or infringe third-party rights.

Where the copyright or licensing status of AI-generated material is uncertain, authors should exercise caution and consult the relevant editorial office. KMF Publishers may request replacement, modification, supporting documentation, or removal of material where intellectual-property concerns arise.

20. Bias, Fairness and Inclusivity

AI outputs may contain historical, cultural, geographic, linguistic, gender-related, racial, disciplinary, or other biases. Authors must critically evaluate AI outputs rather than treating them as neutral or authoritative.

Particular care should be exercised when AI is used to classify people, interpret human behaviour, analyse culturally specific evidence, make predictions, or generate recommendations that could affect individuals or communities.

Human review must consider both technical accuracy and potential unfairness or exclusion.

21. AI Use by Peer Reviewers

Submitted manuscripts are confidential scholarly documents. Reviewers must not upload an unpublished manuscript, identifiable manuscript content, figures, tables, supplementary files, author correspondence, or confidential editorial information into public or unauthorized third-party AI systems.

AI must not replace the reviewer’s independent scholarly judgment. Limited AI assistance for improving the language, grammar, or organization of a review may be acceptable where confidentiality is fully protected. Reviewers remain entirely responsible for their review and may be required to disclose substantive AI assistance.

Any AI use that compromises manuscript confidentiality may result in cancellation of the review and removal of the reviewer from the review process.

22. AI Use by Editors

Editors must preserve confidentiality, editorial independence, and human accountability. Confidential manuscripts must not be uploaded into unauthorized external AI systems.

AI may assist authorized administrative or technical processes where appropriate safeguards exist, but it must not independently make final decisions regarding acceptance, rejection, research integrity, authorship disputes, ethical allegations, or other decisions requiring editorial judgment.

Final editorial responsibility must remain human.

23. AI Use by KMF Publishers

KMF Publishers recognizes that responsible AI governance must apply not only to authors and reviewers but also to the publisher itself.

Where KMF Publishers uses AI-enabled systems in editorial, production, discovery, integrity screening, metadata, language, or administrative workflows, appropriate human oversight, confidentiality, data protection, fairness, and accountability should be maintained.

KMF Publishers will not intentionally use confidential submitted manuscripts to train public generative AI models without appropriate legal authority or permission.

Automated systems may support editorial workflows, but material scholarly and ethical decisions remain subject to human oversight. This principle reflects symmetrical accountability: responsible AI obligations apply across the scholarly publishing ecosystem.

24. AI Detection and Due Process

KMF Publishers does not regard an AI-detection score, by itself, as conclusive evidence of misconduct.

AI-detection technologies may produce uncertain or erroneous classifications and should therefore be treated only as possible screening or supporting information.

No manuscript should be rejected, and no author should be accused of misconduct, solely because an automated detector labels text as AI-generated.

Where concerns arise, editors should consider the manuscript, disclosures, supporting evidence, author explanation, research records where appropriate, and other relevant information. Authors must be given a reasonable opportunity to respond to material allegations.

Human judgment and procedural fairness must prevail over automated accusation.

25. KMF AI Provenance Record

For substantial Level II or Level III AI use, authors are encouraged—and may be required—to maintain an AI Provenance Record.

The record should document, as relevant, the AI system used, version, date of use, purpose, manuscript or research component affected, nature of human intervention, verification performed, and important prompts or workflow information necessary for research integrity or reproducibility.

The record does not normally need to be published in full but should be retained by the authors and supplied to KMF Publishers where reasonably required for editorial or research-integrity assessment.

This creates an auditable chain of human–AI scholarly responsibility.

26. Mandatory AI Disclosure Statement

Where disclosure is required, authors should include a section before the References entitled “Declaration of Generative AI and AI-Assisted Technologies.”

Standard wording: “During the preparation and/or research of this work, the author(s) used [NAME AND VERSION OF AI TOOL] for [PURPOSE]. The AI-assisted output was critically reviewed, verified, and edited by the author(s). The author(s) independently verified the accuracy of relevant facts, data, citations, references, and interpretations and take full responsibility for the integrity and content of this publication.”

Where AI was integral to the research methodology, additional methodological details must be provided in the Methods/Methodology section.

Where no declarable AI was used, authors may state: “The author(s) declare that no generative AI or AI-assisted technology requiring disclosure under the KMF Publishers AI Policy was used in the preparation or research of this manuscript.”

27. Undisclosed or Inappropriate AI Use

Failure to disclose material AI use may constitute a breach of publication transparency. The seriousness of the breach will be assessed proportionately, taking into account the nature of the AI use, its effect on the scholarly record, whether deception was involved, and whether the integrity of data, results, authorship, citations, or conclusions was compromised.

Before publication, appropriate action may include clarification, request for disclosure, revision, additional review, rejection, or investigation.

After publication, appropriate action may include an editorial notice, correction, expression of concern, retraction, or other action consistent with accepted publication-ethics principles.

KMF Publishers distinguishes between responsible disclosed AI assistance and deceptive AI use.

28. Appeals and Fair Treatment

Authors have the right to respond to concerns regarding alleged inappropriate AI use. KMF Publishers will seek to evaluate such concerns fairly, confidentially, consistently, and on the basis of available evidence.

The purpose of this policy is not to punish legitimate technological assistance. Its purpose is to protect transparency, scholarly originality, reproducibility, accountability, and trust.

29. Post-Publication Responsibility

AI-related responsibility does not end at publication.

If authors discover that AI-assisted material introduced a material factual error, fabricated reference, inaccurate quotation, problematic image, analytical error, or other issue affecting the scholarly record, they should promptly inform KMF Publishers.

KMF Publishers will evaluate whether clarification, correction, retraction, or another post-publication action is appropriate.

30. Dynamic Policy and Future Technologies

Artificial intelligence is developing rapidly. Accordingly, this is a living policy.

KMF Publishers may revise these guidelines in response to technological developments, international publication-ethics standards, legal requirements, research-community expectations, or new forms of AI-assisted scholarship.

Where a new technology is not explicitly covered, the underlying principles of this policy remain applicable: Human Intellectual Leadership → Transparency → Verification → Confidentiality → Reproducibility → Fairness → Accountability.

31. KMF Responsible AI Standard

AI may assist-but must not deceive.

AI may accelerate-but must not fabricate.

AI may recommend-but humans must evaluate.

AI may generate-but authors must verify.

AI may support research-but must not replace scholarly judgment.

AI may contribute to the workflow-but cannot assume authorship or accountability.

Confidential scholarship must remain confidential.

Automated detection must never replace evidence and due process.

The more substantial the AI contribution, the greater the requirement for transparency and reproducibility.

Responsibility for scholarly work ultimately remains human.

32. Editorial Discretion

Individual KMF journals, books, disciplines, or publication programmes may adopt additional requirements where necessary because of disciplinary standards, research ethics, legal obligations, or the nature of the material being published.

Where discipline-specific requirements are stricter than this general policy, the stricter applicable requirement should normally be followed.

Authors uncertain about a particular use of AI are encouraged to contact the relevant editorial office before submission.

33. Policy Commitment

KMF Publishers supports innovation in scholarly communication while maintaining the fundamental principles upon which academic publishing depends: originality, transparency, methodological rigor, intellectual independence, confidentiality, reproducibility, fairness, and human accountability.

KMF Publishers does not seek a future in which artificial intelligence replaces researchers, authors, reviewers, or editors.

We support a future in which responsible AI strengthens human scholarship while the authority, judgment, creativity, and responsibility of scholarly communication remain fundamentally human.

International Reference Framework

This policy has been developed with reference to evolving international practices in scholarly publishing. The following external policies and guidance were consulted as comparative reference points. KMF Publishers retains full responsibility for the content, interpretation, implementation, and future revision of its own institutional policy.

Oxford University Press – Policy on AI Use & Disclosure for Oxford Journals Authors

Springer Nature – AI Use in Manuscript Preparation

Taylor & Francis – Using Artificial Intelligence (AI) in Research and Manuscript Preparations

Elsevier – Generative AI Policies for Journals

Wiley – AI Guidelines for Authors

Sage – Generative AI Guidelines for Book and Learning Resource Authors

MDPI – Research and Publication Ethics

Publisher Information

KMF Publishers
https://kmf-publishers.com/
Policy Version: 1.0
Effective: September 2026
Review Cycle: Regular review in response to developments in AI and international scholarly-publishing standards.

© 2026 KMF Publishers. All rights reserved.