Use of Artificial Intelligence

Last Updated: October 2026

Purpose and Scope

ETFLIN recognizes that artificial intelligence (AI) can support researchers in preparing manuscripts, developing research ideas, improving scientific communication, and performing selected technical tasks. However, its use must not compromise research integrity, scientific accuracy, originality, intellectual property rights, confidentiality, or the accountability of authors and editorial personnel.

This policy establishes the conditions under which AI-assisted technologies may be used in the preparation, conduct, evaluation, and publication of scholarly work. It applies to authors, co-authors, reviewers, editors, editorial board members, and other individuals involved in the publishing process, as well as to AI-assisted activities undertaken by ETFLIN where relevant.

For the purposes of this policy, AI-assisted technologies include generative AI systems, large language models, chatbots, AI-based translation and writing tools, AI-assisted coding systems, and image-generation applications. The policy also applies to other AI-enabled tools when their use materially contributes to a study, its analysis, the preparation of a manuscript, or the editorial process.

The fundamental principle is that AI may assist human work but must not replace human responsibility. The use of AI does not diminish the obligations of researchers to conduct sound research, provide reliable evidence, acknowledge relevant sources, comply with applicable ethical requirements, and present their findings honestly.

Where individual journals establish additional requirements for particular disciplines, article types, or research methods, authors and editorial personnel must comply with those requirements in addition to this policy.

AI and Authorship

AI systems, chatbots, language models, and other AI-assisted technologies must not be listed as authors or co-authors of a submitted or published manuscript. They cannot independently accept responsibility for research, approve the final manuscript, disclose competing interests, or fulfill the ethical and legal obligations associated with authorship.

Authorship is reserved for human contributors who satisfy the journal's authorship criteria and can take responsibility for their contributions to the work. AI tools must not be included in the author byline, author contribution statement, or author list presented in the manuscript submission system.

The use of AI must not be presented in a manner that misrepresents the intellectual contributions of the human authors. Authors must not use AI-generated text, analysis, or other material to create a misleading impression of scientific expertise, independent verification, original experimentation, or personal contribution.

Any substantial contribution made by AI must instead be described transparently in the appropriate disclosure section. Such disclosure does not confer authorship or transfer responsibility away from the human authors.

Permitted Uses of AI by Authors

We permits the responsible use of AI-assisted technologies to support manuscript preparation and selected research-related tasks, provided that their use complies with this policy, the journal's subject-specific requirements, and applicable legal and ethical standards.

Acceptable uses include improving the readability, clarity, structure, and organization of author-written manuscripts; correcting grammar, spelling, punctuation, and sentence construction; translating text between languages; and assisting authors in organizing existing content. Authors may also use AI to develop preliminary ideas, explore possible research questions, clarify concepts, or consider alternative ways of presenting an argument. Such assistance must not replace independent scholarly judgment or be used to misrepresent the origin of intellectual contributions.

AI may be used to assist with coding, including code review, debugging, identification of potential programming errors, and suggestions for improving code efficiency. Authors remain responsible for testing the code, evaluating its suitability, validating its outputs, and ensuring that the computational procedures are reproducible. When AI contributes to the research workflow or the generation of scientific results, the relevant procedures must be described in the Methods section.

AI may also assist in preparing tables from verified source data, improving table organization, or converting author-verified information into a clearer format. It must not be used to invent observations, fabricate numerical values, alter results to support a hypothesis, or generate unsupported interpretations. Every table must accurately represent the underlying data, calculations, and methods.

AI-generated illustrative material is permitted only within the restrictions described in the section on figures and visual content. Permitted uses include explanatory diagrams, conceptual schematics, and workflow illustrations that communicate ideas or procedures rather than purporting to present experimental evidence.

These permissions do not authorize the submission of unverified AI output as established scientific knowledge. Authors must independently evaluate all content that contributes to the submitted work, regardless of whether it was generated, translated, summarized, reorganized, or technically reviewed by AI.

Prohibited and Restricted Uses of AI

AI must not be used to fabricate, falsify, or deceptively modify research data, experimental observations, statistical results, references, source material, or scientific conclusions. The generation of plausible-looking but nonexistent evidence is unacceptable, even when the resulting material appears scientifically credible.

AI-generated content must not be used to conceal plagiarism, misappropriate another person's ideas, reproduce protected material without appropriate authorization, or create a misleading impression of originality. Authors must not present AI-generated claims as independently verified facts or cite references solely because an AI system has suggested them.

The use of AI must not replace research activities that require actual observation, measurement, experimentation, or analysis. Where an AI system is an essential component of the research method itself, its use is subject to appropriate methodological reporting, validation, and ethical oversight. The system's involvement must not be concealed by describing AI-generated outputs as direct observations or results obtained through procedures that were not actually performed.

AI must not be used to generate figures, images, or other visual materials that purport to represent actual research evidence, experimental observations, clinical findings, microscopy, histology, laboratory measurements, or other primary data. AI must not be used to create or alter scientific images in ways that introduce, remove, obscure, or misrepresent evidence. Genuine data visualizations must be generated from verified underlying data using appropriate, reproducible methods rather than fabricated through image generation.

The creation of graphical abstracts using generative AI is prohibited. Graphical abstracts must be prepared using appropriate non-generative design or scientific illustration tools and must accurately represent the work described in the manuscript. All components must be original or used with appropriate authorization.

AI-generated illustrative diagrams, conceptual schematics, and workflow illustrations may be considered only when their purpose is explanatory, their content can be verified, and they do not misrepresent the study or its findings. Their use must be disclosed as specified in this policy.

We may request original files, source data, code, prompts, or other supporting documentation when necessary to assess the authenticity, accuracy, and integrity of material submitted for publication.

Disclosure of AI Use by Authors

Transparency is a central requirement of responsible AI use. Authors must disclose AI-assisted technologies that contribute to the conduct of a study, the analysis of data, the preparation of a manuscript, or the creation of material submitted for publication, subject to the limited exception for routine language correction described below.

AI use must be declared during manuscript submission. The corresponding author must answer the submission system's AI-use questions accurately and must not report that no AI was used when AI-assisted technology contributed to the study or manuscript. The disclosure requirement applies regardless of whether the AI-generated material appears in the final manuscript, provided that the use materially contributed to the work being submitted.

A submission-stage declaration does not automatically replace the disclosure required in the manuscript itself. Where substantive or methodological AI use has occurred, the relevant information must also be included in the appropriate manuscript section. We may request further details when the submitted declaration does not adequately describe the use.

The exact location of the manuscript disclosure depends on the nature of the activity. AI used as part of the research design, data collection, computational workflow, data analysis, code development, or other reproducible procedure must be described in the Methods section. The description must be sufficiently detailed to allow readers and reviewers to understand the tool's role and evaluate or reproduce the relevant procedure.

AI used for writing assistance, substantial language revision, content organization, translation, or the development of manuscript text must be acknowledged in the Acknowledgments section or in the journal's designated AI disclosure statement. If the use affects both the research methods and the written presentation, disclosure may be required in both sections. The same disclosure principles apply to permitted AI-generated illustrative material, which must additionally be identified in the relevant figure caption.

Routine spelling, punctuation, grammar correction, and minor language polishing that do not materially change the meaning, scientific content, interpretation, or intellectual contribution of the manuscript may be exempt from a separate disclosure statement in the published article. Basic translation or wording assistance may receive the same treatment only when it is genuinely limited to routine language support. Substantive translation, extensive rewriting, generation of new content, or AI-assisted changes to the meaning or interpretation of the work must be disclosed. Authors must follow any more specific disclosure requirements established by the relevant journal.

The absence of a published disclosure requirement for routine language correction does not permit authors to provide a false answer to an explicit question in the submission system. Where the journal requests information about any AI use, authors must report that use accurately.

If no AI-assisted technology has been used in the study, research workflow, or preparation of the manuscript, no AI-use disclosure statement is required. Authors should not insert a statement claiming that AI was used when it was not.

Required Information in an AI Disclosure

An AI-use declaration must provide enough information for the Publisher, reviewers, editors, and readers to understand the nature and extent of the assistance received.

The declaration must identify the name of the AI tool or service, the model used, the model or software version when available, and the developer or provider when relevant. It must explain the purpose of the use, the stage of the research or manuscript preparation at which it occurred, and the scope of the material or activity affected. Authors must also describe the nature of the human review, verification, and revision undertaken after the AI-generated output was received.

When AI contributes to a research method, data analysis, code development, or the generation of results, the Methods section must provide sufficient technical detail to permit appropriate scrutiny and reproducibility. This includes the model or system configuration and relevant settings when these materially affect the procedure. For analyses that depend on specific prompts or instructions, authors should provide the prompts, relevant inputs, and other necessary procedural details in the Methods section or supplementary materials, as appropriate.

Authors are encouraged to retain and provide the prompts used to generate substantive text, tables, or permitted illustrations when those prompts are relevant to evaluating the output or reproducing the process. This is particularly important when AI forms part of the methodology or materially influences scientific content. Prompt disclosure must be handled in a way that does not expose confidential information, personal data, protected information, or third-party material without authorization.

For minor, routine language correction that qualifies for the disclosure exception, a full record of individual prompts is not normally required unless requested by the journal. When we requests clarification, authors must provide an accurate account of the use to the extent reasonably available.

Authors must not use a generic statement such as "AI was used to improve the manuscript" when the tool also contributed to the research design, data analysis, generation of tables, or other substantive activities. The declaration must distinguish among these activities rather than treating all AI assistance as a single undifferentiated process.

Verification, Accuracy, and Accountability

All authors are jointly responsible for the integrity of the submitted work and must understand and approve any AI use that materially contributes to it. The corresponding author must ensure that all co-authors have been informed of the use, have had an appropriate opportunity to review the affected material, and agree to the disclosures made to the journal.

AI-generated content must be carefully checked before it is incorporated into a manuscript. AI systems may produce fabricated references, incorrect quotations, inaccurate translations, misleading summaries, unsupported claims, biased interpretations, or plausible but erroneous technical explanations. Fluent language or a confident presentation must never be treated as evidence of accuracy.

Authors must verify the existence and bibliographic details of every cited reference against reliable source records. They must read the relevant sources and ensure that they genuinely support the claims attributed to them. References suggested by AI must not be accepted without independent confirmation. AI-generated text is not a substitute for consulting original research, authoritative documents, or other appropriate primary sources.

Authors must independently evaluate the originality, relevance, accuracy, and scientific appropriateness of all AI-assisted content. They must check whether the output introduces unsupported claims, omits important qualifications, changes the meaning of the original work, or reproduces material from existing publications. Where the manuscript presents quantitative findings, authors must verify the calculations, data transformations, code, and reported results using appropriate methods.

All authors must ensure that their manuscript accurately reflects the research actually conducted and the evidence available. Responsibility for data integrity, proper attribution, research ethics, and scientific conclusions remains with the authors, regardless of the level of AI assistance. AI use must not be offered as a justification for errors, omissions, misrepresentation, or breaches of publication ethics.

Before submission, every author must be aware of the applicable AI disclosure, must agree to the use of AI described in the manuscript, and must be satisfied that the final work meets the journal's standards. The corresponding author must not make AI-related declarations on behalf of co-authors without ensuring that the declarations are accurate and agreed upon.

Copyright, Intellectual Property, and Originality

Authors are responsible for ensuring that all content submitted for publication complies with applicable copyright law, licensing conditions, intellectual property rights, and the terms governing the AI tools used.

The fact that content has been generated by AI does not automatically establish that it is original, free from third-party rights, or eligible for unrestricted publication. AI-generated text, illustrations, tables, and other materials must be examined for potentially reproduced or substantially similar protected content. Authors must not assume that an AI service grants all rights required for publication or that all generated outputs are copyright-free.

Before submitting AI-assisted material, authors must review the applicable tool or service terms, confirm that the material can lawfully be used for the intended publication, and obtain any permissions required for third-party content. Where licensing or ownership is uncertain, authors must resolve the issue before submission or replace the material with content for which appropriate rights can be established.

AI tools must not be used to circumvent copyright restrictions, remove attribution, imitate protected works in an unauthorized manner, or reproduce another person's distinctive material without appropriate legal grounds. Any third-party material incorporated into an AI-assisted output remains subject to applicable attribution and permission requirements.

Authors must also ensure that AI-assisted writing does not introduce plagiarism or obscure the contribution of existing scholarship. Original sources must be acknowledged appropriately, and quotations must be attributed to the correct source. Authors must not cite an AI system as the original scientific authority for a factual claim when an appropriate underlying source is available.

Where necessary, we may request information about the source of an illustration, the terms of use of an AI service, relevant permissions, or other evidence required to establish that content can legitimately be published.

AI-Generated Images, Figures, and Visual Content

ETFLIN distinguishes explanatory illustrations from images that constitute research evidence. This distinction is necessary because a conceptual illustration communicates an idea, whereas a research image or figure may be used to substantiate an observation, measurement, or scientific conclusion.

Generative AI may be used only for permitted explanatory illustrations, such as conceptual diagrams, workflow diagrams, process schematics, and other non-evidentiary visual explanations. Such material must be scientifically appropriate, free from misleading representations, and independently reviewed by the authors. Any AI-assisted illustration must be clearly disclosed in the figure caption and in the manuscript's general AI disclosure statement.

The caption must identify the tool and model or version when available, explain the role of AI in producing the illustration, and indicate whether the image was subsequently edited or substantially modified by the authors. Authors must ensure that the final illustration accurately reflects the concepts and processes it is intended to communicate.

AI must not be used to generate or modify images presented as direct evidence from the research, including laboratory images, clinical images, microscopy, histology, gels, blots, patient images, or other primary experimental material. It must not be used to add, remove, synthesize, or modify visual features in a manner that could affect scientific interpretation.

Graphical abstracts must not be created using generative AI, even when their purpose is illustrative. Authors must prepare graphical abstracts using suitable non-generative design or scientific illustration tools and verify that every component is accurate, original, and appropriately licensed.

AI-assisted creation of a permitted illustration does not remove the authors' obligations concerning scientific accuracy, attribution, accessibility, copyright, or permissions. ETFLIN retains the right to reject visual material that is inaccurate, misleading, improperly licensed, inadequately disclosed, or inconsistent with the journal's editorial requirements.

AI Use by Reviewers

Peer reviewers are responsible for providing independent, critical, fair, and evidence-based assessments of submitted manuscripts. The use of generative AI to evaluate scientific merit, assess the validity of methods, formulate substantive criticisms, determine whether conclusions are justified, or generate the main content of a peer-review report is strongly discouraged and is not permitted under this policy.

Reviewing requires disciplinary expertise, critical reasoning, contextual judgment, and accountability for the assessment provided. These responsibilities must not be delegated to an AI system. Reviewers must personally evaluate the manuscript and formulate their scientific judgments without relying on AI-generated evaluations as a substitute for their own assessment.

Reviewers must not use AI to decide whether a manuscript should be accepted, rejected, or revised. Any recommendation concerning publication must arise from the reviewer's own assessment of the evidence, methods, originality, relevance, and scientific contribution. Reviewers remain accountable for every criticism, recommendation, and statement included in their reports.

Limited language assistance may be permitted to improve the grammar, clarity, or tone of a reviewer's own comments, provided that it does not change the substance of the assessment, introduce new criticisms, generate a scientific judgment, or compromise confidentiality. Such assistance must be undertaken only through an ETFLIN-approved environment with appropriate confidentiality protections. Reviewers must not upload a submitted manuscript, its excerpts, figures, supplementary files, review questionnaires, confidential correspondence, or substantive review comments to public or otherwise unapproved AI services.

If a journal requires disclosure of AI use by reviewers, the reviewer must provide an accurate declaration to the editor in accordance with that journal's instructions. Any permitted use must not diminish the reviewer's responsibility for the accuracy, fairness, and appropriateness of the final report.

AI Use by Editors and Editorial Staff

Editors are responsible for ensuring that editorial decisions are made independently, fairly, and in accordance with the journal's editorial standards. Generative AI must not be used to make, determine, or independently recommend editorial decisions, including decisions on acceptance, rejection, revision, reviewer selection, or the resolution of research-integrity concerns.

AI must not replace an editor's independent evaluation of the manuscript, the reviewers' reports, the validity of research claims, or the ethical implications of a submission. Editors must make editorial decisions based on their own professional judgment, the evidence available, applicable publication policies, and advice from qualified human experts when necessary.

AI may be used for limited administrative or language-support tasks where expressly permitted by the ETFLIN, such as improving the grammar and readability of an editor's own correspondence. These activities must not influence substantive judgments or introduce unverified claims into communications with authors or reviewers.

Any permitted language assistance involving editorial communications must take place in a secure, ETFLIN-approved environment. Editors must not submit confidential manuscript content, reviewer reports, author correspondence, editorial deliberations, or unpublished research to public or unapproved AI services. The confidentiality obligations of editorial personnel remain in force even when a tool is used solely to improve language.

Editors must not treat AI-generated assessments, AI-detection scores, or automated indications of possible misconduct as conclusive evidence. Where concerns arise regarding originality, authorship, fabricated content, or research integrity, editors must investigate them using appropriate evidence, established editorial procedures, and fair opportunities for the relevant parties to respond.

We may use AI for limited internal support functions under appropriate governance, security, and human oversight. Such use must not undermine confidentiality or transfer editorial authority to an automated system. Where ETFLIN's use of AI materially affects the editorial process or the handling of a manuscript, the use must be documented and disclosed where required by the journal's procedures or applicable standards.

Confidentiality, Privacy, and Data Protection

Submitted manuscripts, unpublished research findings, reviewer reports, editorial correspondence, and related submission materials are confidential. They must be handled only for the purposes for which they were provided and must not be disclosed to unauthorized individuals, organizations, or AI systems.

Authors, reviewers, editors, and editorial staff must not upload confidential manuscripts or associated materials to public or unapproved AI services. This prohibition applies to entire documents as well as excerpts, abstracts, figures, tables, datasets, supplementary information, peer-review comments, and prompts containing confidential information. Removing an author's name or other obvious identifiers does not necessarily eliminate confidentiality risks.

Before using an AI tool for a permitted task involving sensitive information, the user must verify that the tool has been approved for that purpose and that appropriate safeguards are in place. Relevant considerations include data retention, reuse for model training, access by third parties, security controls, data storage, and compliance with applicable privacy and contractual obligations. A service must not be assumed to be secure merely because it is described as private or confidential.

Personal data, patient information, identifiable research records, confidential institutional information, and proprietary datasets must not be submitted to AI services without an appropriate legal basis, authorization, and safeguards. Where necessary, users must remove or sufficiently de-identify sensitive information before processing it, while ensuring that doing so does not create other ethical or methodological problems.

Reviewers must not share manuscript content with colleagues or external services without the editor's authorization and must not use AI to circumvent the confidentiality requirements of peer review. Editors and ETFLIN as a publisher must ensure that any approved technology used in editorial workflows has appropriate security, access controls, and data-handling arrangements.

Any suspected disclosure, unauthorized transfer, or misuse of confidential material through an AI system must be reported promptly to the appropriate editorial or publishing contact. We will assess the incident and take proportionate action to protect affected individuals, unpublished work, and the integrity of the editorial process.

Undisclosed or Inappropriate AI Use

Failure to comply with this policy may raise concerns about research integrity, transparency, authorship, copyright, or publication ethics. An omission of an AI disclosure does not automatically establish misconduct; however, intentional concealment, false declarations, or the use of AI to fabricate evidence or misrepresent scientific contributions may constitute serious breaches of publication ethics.

When a potential violation is identified, we (ETFLIN) or journal editor may request clarification, additional disclosures, original data, source files, code, prompts, permissions, or other relevant documentation. The affected parties must be given an appropriate opportunity to explain their use of AI and respond to the concern.

Depending on the nature and severity of the issue, and subject to the journal's established procedures, possible actions may include a request for correction of the disclosure, revision of the manuscript, removal or replacement of material that does not meet the policy, rejection of the submission, or further action under the ETFLIN's research-integrity and publication-ethics procedures. For published articles, the response may include a correction, an expression of concern, or retraction where warranted by the evidence and applicable procedures.

We will consider the circumstances of each case, including whether the use was disclosed, whether the affected content is reliable, whether rights or confidentiality were compromised, and whether the issue can be addressed without undermining the validity of the publication. Any decision must be made by authorized human editorial personnel, not by an AI system.

Responsibilities of ETFLIN

ETFLIN is responsible for communicating this policy clearly to authors, reviewers, editors, and editorial staff and for making the relevant instructions available through journal websites, submission systems, reviewer guidance, and editorial procedures.

We will establish appropriate processes for collecting AI-use declarations, assessing potential non-compliance, protecting confidential information, and maintaining consistency across its journals. Editors and staff should receive sufficient guidance to distinguish permitted assistance from prohibited use and to manage AI-related concerns fairly and responsibly.

Where AI tools are used by the Publisher in manuscript processing, language editing, production, or other publishing activities, their use must be governed by appropriate standards of human oversight, security, accuracy, and transparency. AI-generated or AI-assisted content published under the Publisher's name must receive suitable human review and must not be presented as verified without appropriate checks.

We will not treat AI-detection software as definitive proof that a manuscript was generated by AI or that misconduct has occurred. Any concern must be assessed in context, using multiple forms of relevant evidence and established editorial procedures. Authors must not be subjected to adverse editorial action solely because an automated tool has flagged their work.

We may review and update this policy as AI technologies, legal requirements, scholarly practices, and publication-ethics standards evolve. Updates will be communicated through the appropriate publishing channels. All parties are responsible for consulting the current version of this policy and any additional instructions issued by the journal to which a manuscript is submitted.

Final Principle

AI may serve as a supporting instrument in scholarly research and communication, but scientific integrity remains a human responsibility. Authors must ensure the accuracy and originality of their work; reviewers must provide independent assessments; editors must exercise independent editorial judgment; and the Publisher must safeguard confidentiality, transparency, and the reliability of the scholarly record.

The acceptability of AI use is determined not merely by the tool employed, but by the purpose of its use, the reliability of its output, the transparency of its contribution, the protection of intellectual property and confidential information, and the quality of human oversight. Any use that compromises these principles is inconsistent with this policy.