Summary
The International Committee of Medical Journal Editors (ICMJE) continues to shape global standards for ethical medical publishing. Its 2025 updates respond to major developments in research practice, especially the rise of AI-assisted writing, increasing expectations for data transparency, growing concern about predatory journals, and the need to clarify authorship and peer-review responsibilities.
The revised guidance tightens the definition of authorship and reinforces the requirement that all listed authors make substantial intellectual contributions, approve the final manuscript, and accept accountability for the work. Crucially, AI tools cannot be credited as authors. Any use of AI in drafting, editing, translation, image generation, or data analysis must be fully disclosed, and all AI-generated content remains the responsibility of human authors, who must verify facts, references, and interpretations.
The 2025 updates also strengthen expectations for data sharing through mandatory data-availability statements and clearer policies on corrections and retractions when errors are discovered. To combat predatory publishing and duplicate submissions, ICMJE urges authors to verify journal credibility, avoid simultaneous submissions, and respect restrictions on redundant publication and self-plagiarism. Journals are encouraged to use plagiarism-detection tools systematically and to educate authors about publication ethics.
Peer review receives renewed focus, with recommendations for reviewer training, conflict-of-interest disclosure, blinded or open review options, and better monitoring of reviewer conduct. For researchers, these updates mean more detailed disclosures, stricter authorship scrutiny, and closer attention to data management and journal selection. For editors and publishers, they require clearer policies, stronger screening systems, and transparent communication with authors. In this environment, relying solely on AI text generation or correction is risky: carefully prepared manuscripts, supported where needed by expert human academic proofreading, remain the safest route to compliant, credible medical publications.
📖 Full Length Article (Click to collapse)
ICMJE 2025 Updates: How New Guidelines Shape Ethical Medical Publishing
Introduction
The International Committee of Medical Journal Editors (ICMJE) exerts enormous influence over how medical and biomedical research is reported worldwide. Its recommendations underpin the policies of thousands of journals, shape institutional guidance, and affect how careers are built and evaluated. As research practices, technologies, and ethical challenges evolve, the ICMJE periodically revises its guidance to keep pace.
The 2025 updates are the latest step in this process. They respond to a rapidly changing publishing environment in which AI-assisted writing tools are increasingly used, expectations for data transparency and reproducibility are higher than ever, and concerns about predatory journals, duplicate publications, and peer-review integrity continue to grow. At the heart of the revisions is a simple goal: to reinforce transparency, accountability, and trust across the entire research lifecycle.
This article explains the key areas targeted by the ICMJE 2025 recommendations, focusing on five pillars of responsible publishing:
- Authorship responsibility and ethical conduct
- AI and its role in manuscript preparation
- Data sharing and transparency
- Predatory journals, plagiarism, and duplicate publications
- Peer-review accountability and integrity
We also explore what these updates mean in practice for researchers, editors, and publishers, and why careful, human-led manuscript preparation—potentially supported by professional academic proofreading—has become more important than ever in an era of AI-generated text and tightening scrutiny.
Key Areas Addressed in the ICMJE 2025 Recommendations
The 2025 revisions do not replace the entire ICMJE framework; instead, they strengthen and clarify specific areas where new challenges have arisen. The five focal themes are:
- Authorship responsibility and ethical conduct
- AI and its role in manuscript preparation
- Data sharing and research transparency
- Predatory publishing and duplicate submissions
- Peer-review accountability and professional standards
Each of these themes reflects both long-standing ethical concerns and newer issues brought about by digital technologies, AI, and rapidly expanding publication volumes.
1. Authorship Responsibility and Ethical Conduct
Authorship is central to academic recognition and responsibility. The ICMJE has long used its four-part authorship criteria to distinguish true authors from contributors, collaborators, or sponsors. In 2025, this framework is reaffirmed and further emphasised.
Refining Authorship Criteria
To qualify as an author, a contributor must:
- Make substantial contributions to the conception or design of the work, or to the acquisition, analysis, or interpretation of data.
- Participate actively in drafting the manuscript or revising it critically for important intellectual content.
- Approve the final version of the manuscript before submission.
- Agree to be accountable for the integrity and accuracy of the work, answering questions related to their part of the research and cooperating in investigations of potential errors or misconduct.
The 2025 updates reiterate that all four criteria must be met. Individuals who contribute only data collection, funding acquisition, general supervision, or administrative support should be recognised in the acknowledgements but not listed as authors.
Clarifying the Role of AI in Authorship
One of the most visible changes in 2025 is the explicit statement that AI tools cannot be named as authors. Although AI systems can assist in drafting text, translating content, or summarising literature, they cannot fulfil the core criteria of responsibility and accountability. They cannot agree to be accountable, respond to critiques, or correct the record.
Therefore, even when AI is used extensively, the human authors remain fully responsible for the accuracy, originality, and ethical integrity of the submitted work. They must verify all AI-generated text, check references for accuracy, and ensure that language is not misleading or plagiarised. This is one reason why many researchers choose to complement AI tools with expert human proofreading services to ensure that final manuscripts meet ethical and stylistic expectations.
Preventing Ghost and Honorary Authorship
The ICMJE also renews its warnings against ghost authorship (unacknowledged writers or analytic contributors) and honorary authorship (listed authors who did not meet the criteria). Journals are encouraged to require detailed contributorship statements describing each author’s role—such as conceptualisation, methodology, data curation, writing, or supervision.
This level of transparency makes it harder to add senior names simply for prestige or to hide the involvement of commercial editing agencies or undisclosed AI-driven writing services.
2. AI and Its Role in Manuscript Preparation
AI has rapidly become part of the manuscript-preparation toolkit, but it introduces serious questions related to originality, accuracy, and accountability. The 2025 recommendations therefore address AI use head on.
Disclosure of AI Use
ICMJE now expects authors to explicitly disclose any significant use of AI tools in manuscript preparation. This includes AI involvement in:
- Drafting or rewriting sections of the text.
- Translating content between languages.
- Generating or editing images, diagrams, or visual abstracts.
- Summarising literature, generating search strategies, or proposing analyses.
This disclosure typically belongs in the Methods or Acknowledgements section and should specify which tools were used and for what purpose. Hidden AI use is discouraged and may be treated as a form of misrepresentation.
Verifying AI-Generated Content and References
Because AI systems can produce fabricated citations, incorrect facts, or distorted summaries, the ICMJE emphasises that all content produced with AI support must be independently verified by human authors. This includes checking:
- That all references are real, relevant, and accurately cited.
- That summaries correctly represent the original research findings.
- That no confidential or sensitive data were inadvertently uploaded to external AI services.
AI may be helpful at early drafting or idea-generation stages, but final manuscripts must be human-validated. Overreliance on AI for rewriting without understanding the material can easily produce subtle plagiarism or misinterpretation of results.
Prohibiting AI in Data and Image Manipulation
The recommendations also stress that AI must not be used to fabricate or manipulate data, images, or results. AI-generated figures or images must be clearly labelled, and any enhancement should be scientifically justified and reproducible. Using AI to “clean up” data in a way that alters their meaning is strictly incompatible with good research practice.
3. Data Sharing and Transparency in Research
Transparent data practices are essential for reproducibility and public trust. The ICMJE 2025 update strengthens expectations around data-sharing and documentation.
Mandatory Data-Availability Statements
Most clinical trials and major research studies are now expected to include a data-availability statement (DAS) specifying:
- Whether the underlying dataset is publicly available, available on request, or restricted.
- Where data can be accessed (for example, in a recognised repository).
- Any conditions or approvals required for data access, particularly for sensitive or patient-level information.
These statements help readers understand how easily findings can be replicated, and they support broader open-science initiatives.
Correcting the Record When Errors Are Found
The ICMJE reiterates that when errors, inconsistencies, or ethical concerns are identified after publication, authors and journals must act quickly. Depending on severity, appropriate actions include:
- Publishing a correction or erratum.
- Issuing an expression of concern while investigations are ongoing.
- Retraction of the article, accompanied by a transparent explanation, in cases of major error or misconduct.
Data should remain available, as far as ethically permissible, to allow independent verification and to support investigations into potential problems.
4. Addressing Predatory Journals and Duplicate Publications
The 2025 guidelines sharpen language around predatory publishing—journals that present themselves as legitimate but lack robust peer review, editorial oversight, or ethical safeguards.
Identifying and Avoiding Predatory Journals
ICMJE encourages researchers to assess journals systematically before submission by checking:
- Whether the journal is indexed in reputable databases or recognised by major libraries.
- The clarity and credibility of its editorial board and peer-review process.
- The transparency of fees and publication timelines.
Institutions are urged to provide training on recognising predatory publishers and to avoid rewarding publications in such outlets when evaluating promotions and funding applications.
Preventing Duplicate Submissions and Redundant Publication
To maintain a clean and trustworthy literature, the ICMJE restates that:
- Manuscripts must be submitted to only one journal at a time.
- Authors should not publish the same or substantially similar work in multiple places without clear cross-referencing and journal permission.
- Self-plagiarism—reusing large parts of one’s own previous text without citation—must be avoided.
Journals are advised to use plagiarism-detection and similarity-check tools during submission to identify overlap early, and to have clear policies on how to handle redundant publication when it is discovered.
5. Enhancing Peer-Review Accountability
Peer review remains a cornerstone of medical publishing, yet it is not immune to bias, inconsistency, or misuse. The 2025 updates promote a more accountable and professional peer-review culture.
Training and Supporting Reviewers
ICMJE recommends regular training for reviewers on topics such as confidentiality, constructive feedback, handling conflicts of interest, and recognising plagiarism or data irregularities. Journals are encouraged to offer guidance documents, webinars, or mentoring systems for new reviewers.
Improving Transparency and Managing Conflicts of Interest
Editors and reviewers should provide clear disclosures of financial and non-financial conflicts of interest that might influence their judgement. Journals may choose to adopt:
- Blinded peer review to reduce identity-based bias.
- Open peer review, where reviewer reports—or even reviewer identities—are made public.
- Internal monitoring of reviewer performance and behaviour.
Whatever model is chosen, consistent policies and transparent communication with authors help build trust in the review process.
Practical Implications for Researchers
For individual researchers, the ICMJE 2025 recommendations translate into several concrete responsibilities:
- Disclose any AI use in manuscript preparation and verify its outputs carefully.
- Ensure that all listed authors meet the full authorship criteria and that contributions are documented.
- Select journals carefully, avoiding predatory outlets and respecting single-submission policies.
- Provide accurate data-availability statements and be prepared to share underlying data where appropriate.
- Use plagiarism-detection tools responsibly and allow time for careful revisions and citation improvements.
Given these expectations, it is risky to rely on AI alone to “clean up” or generate a manuscript. Many authors find it safer to draft in their own words and then work with professional human proofreaders and editors who understand journal conventions, reference styles, and ethical requirements.
Practical Implications for Journals and Publishers
Journals and publishers also face new and reinforced responsibilities under the 2025 framework:
- Implement clear authorship and contributorship policies, including requirements for role descriptions and AI-use disclosures.
- Strengthen plagiarism and similarity screening at submission and pre-acceptance stages.
- Develop and publish policies on data sharing, corrections, and retractions.
- Screen for duplicate submissions and redundant publications.
- Invest in reviewer training and oversight to enhance the quality and fairness of peer review.
By aligning journal policies closely with ICMJE recommendations, publishers can provide authors with clear expectations and contribute to a more trustworthy, efficient, and transparent global research ecosystem.
Conclusion
The ICMJE 2025 updates mark an important evolution in the ethics and practice of medical publishing. By focusing on authorship responsibility, AI transparency, data sharing, predatory publishing, and peer-review accountability, the new guidance aims to preserve what matters most in research: credibility, reproducibility, and public trust.
For researchers, the message is clear: understand the guidelines, document contributions, disclose AI use, select journals carefully, and treat data with care. For journals and institutions, the challenge is to build systems—technical, procedural, and cultural—that support these values and respond quickly when problems arise.
In an era of AI-generated text and increasingly sophisticated similarity checks, hastily prepared manuscripts or poorly controlled AI outputs can easily lead to ethical concerns, inflated similarity scores, or outright rejection. A more reliable approach is to combine responsible use of technology with thorough, human-centred quality control—from careful drafting and accurate referencing through to expert human proofreading. When authors, editors, and institutions work together under the ICMJE framework, they help ensure that medical publications remain not only technically sound, but also ethically robust and worthy of the trust placed in them by patients, practitioners, and the wider public.
