AJB POLICY ON THE USE AND DISCLOSURE OF ARTIFICIAL INTELLIGENCE (AI) IN MANUSCRIPT PREPARATION, SUBMISSION, AND REVIEW

 

A GUIDE FOR AUTHORS, REVIEWERS AND EDITORS

 

POLICY SUMMARY

This policy establishes the African Journal of Bioethics’ standards for the ethical, transparent, and accountable use of Artificial Intelligence in all stages of manuscript preparation, submission, and peer review. AJB recognizes the potential of AI tools to enhance scholarly work while maintaining rigorous standards of research integrity, authorship accountability, and academic honesty.

 

Effective Date: 23rd October, 2025

Approved by: Editorial Board, African Journal of Bioethics

Prepared by: Editorial Team

Review Cycle: Annual

Version: V1 (October 2025)

 

1. Background and Rationale

The African Journal of Bioethics (AJB) recognizes the transformative role of Artificial Intelligence (AI) in modern research, communication and knowledge production. AI tools such as ChatGPT, Gemini, Claude, Copilot and other generative or analytical AI systems are increasingly being used by authors to refine language, summarize text and generate data visualizations. Others are used to produce preliminary research drafts. Although these tools offer convenience and efficiency, they introduce complex ethical and methodological challenges relating to authorship, originality, research integrity and accountability. This policy aims at establishing clear and comprehensive guidance on the acceptable, transparent and responsible use of AI tools in all manuscripts submitted to AJB. This policy is key in keeping the journal’s core mandate of upholding the highest standards of academic integrity and ethical publishing. The policy aims to ensure that the use of AI enhances rather than undermines scholarly rigor, bioethical principles and trust in the integrity of the scientific record.

 

2. Scope and Applicability

This policy applies to all manuscripts submitted to the African Journal of Bioethics, including but not limited to original research articles, review papers, case studies, commentaries, short communications, editorial and special issues. It covers all stages of manuscript preparation, from conceptualization and writing to data analysis and figure generation. The policy applies to all authors, co-authors, contributors, translators, research assistants and any individual or entity involved in the preparation of a manuscript. It covers both generative AI tools, such as ChatGPT and Gemini and analytical or predictive AI systems used in data processing, modeling, or image generation. The policy also applies to AI-assisted translation or summarization platforms such as DeepL, Bard or Google Translate, and any other emerging tools that perform tasks traditionally associated with human scholarly effort.

 

3. Ethical Foundations and Principles

This policy is grounded in the ethical principles of transparency, accountability, integrity, human oversight, fairness and respect for privacy and confidentiality. AJB views transparency as a fundamental requirement in the reporting of scientific work. Authors must therefore clearly disclose any AI use to enable editors, reviewers and readers to understand the extent of technological assistance in the work. Accountability remains central to the policy and human authors are solely responsible for the accuracy, originality and ethical soundness of all content produced with or without AI assistance. The principle of integrity demands that AI-generated material must never misrepresent, plagiarize, fabricate or manipulate information. Human oversight requires that all intellectual contributions and interpretations must be verified and approved by the authors themselves. Fairness entails the conscious avoidance of algorithmic bias or discriminatory data representations, particularly given the journal’s bioethical focus on justice and equity. Finally, respect for privacy obliges authors and reviewers not to upload or expose confidential, identifiable or unpublished data to AI systems that may store or reuse such information.

 

4. Acceptable Use of Artificial Intelligence

The journal recognizes that AI can serve as a valuable tool in enhancing clarity, structure and efficiency during manuscript preparation. Authors may employ AI technologies for the following purposes, provided all outputs are reviewed and validated for accuracy:

Language and Style Editing

Grammar improvement, spelling correction, sentence restructuring, and stylistic refinement to enhance clarity and readability.

Data Visualization and Statistical Analysis

Creation of figures, graphs, and statistical visualizations, provided the underlying models, methods, and datasets are transparently described and fully replicable. Authors must document all methodological choices and model specifications.

Content Organization and Summarization

AI may be used to organize ideas, outline arguments, or summarize non-confidential text, provided human authors validate and revise all outputs.

Technical Assistance

Support with technical coding, figure formatting, and accessibility compliance.

 

Critical Requirement: Authors must carefully review and validate all AI outputs to ensure factual and conceptual accuracy. The final manuscript must reflect human judgment and intellectual contribution at every stage. All authors retain full responsibility for all content.

 

5. Prohibited or Unethical Use of AI

The African Journal of Bioethics strictly prohibits the following use of AI:

• Manuscript Content Generation: Using AI to generate substantial parts of a manuscript (defined as >25% of any section or any portion of results, hypotheses, or novel arguments) without explicit human oversight, verification, and revision.

 

• Data Fabrication or Manipulation: Using AI to fabricate, invent, or materially alter research data, results, citations, references, or findings.

 

• Visual Data Misrepresentation: Employing AI to manipulate, alter, or enhance images, graphs, or tables in ways that could misrepresent original data, obscure methodology, or distort research findings. This includes unauthorized filters, colour adjustments, or image splicing.

 

• Ethical and Consent Materials: Generating or substantially drafting ethical statements, informed consent forms, data protection documents, or other human subjects-related materials using AI, as these require professional ethical discernment, legal expertise, and contextual understanding.

 

• Confidential Data Entry: Uploading identifiable human participant data, proprietary datasets, unpublished research materials, or institutional information to external AI systems (e.g., ChatGPT, Gemini) that may store, retain, or reuse such information.

 

• Undisclosed Use: Failing to disclose any AI use as detailed in Section 7.

 

Consequences: Failure to comply with these standards constitutes a breach of publication ethics and may result in manuscript rejection, article retraction, institutional notification, or blacklisting from future submissions (minimum 2-year embargo; permanent for repeat offenders).

 

Definition of ‘Substantial Parts’: For clarity, ‘substantial parts’ shall be understood as: (i) more than 25% of any manuscript section; (ii) any single section addressing novel intellectual contributions (e.g., hypotheses, arguments, results interpretation, conclusions); or (iii) any content that presents original analysis, reasoning, or interpretation central to the work’s scholarly contribution. Authors uncertain about whether their AI use qualifies as ‘substantial’ are encouraged to contact the Editorial Office at ajb@africanjournalofbioethics.org.

 

6. Authorship, Responsibility and Accountability

Authorship is a human designation that carries intellectual, ethical and legal responsibility. AI systems do not possess these qualities and thus cannot be listed as authors, co-authors or contributors to any manuscript submitted to AJB. AI systems, algorithms, or tools cannot be listed as authors, co-authors, or contributors regardless of their level of involvement in manuscript production.

All listed authors must meet AJB’s authorship criteria, which are aligned with the recommendations of the International Committee of Medical Journal Editors (ICMJE).  The corresponding author bears primary responsibility for ensuring that all co-authors are informed of any AI use during manuscript preparation and that such use complies with this policy. All authors are jointly accountable for verifying the accuracy, originality and ethical integrity of the manuscript, including any text, figures or analyses produced with AI assistance.

 

7. Disclosure of AI Use

Full and transparent disclosure of AI use is mandatory. Disclosures must be placed as follows:

For Language Editing and Style Refinement: Include a statement in the Acknowledgments section. 

For Data Analysis, Visualization, or Statistical Methods: Include a statement in the Methods section. 

For All Other Uses: Include a dedicated “AI Use Disclosure” subsection within the Methods section or as a numbered footnote on the title page.

 

All disclosure statements must specify:

1. the name and version of the AI tool, 

2. the company or organization that developed it,

3. the specific purpose(s) for which it was used 

4. a confirmation that the authors reviewed and verified all outputs; and 

5. a statement that authors accept full responsibility for the final content.

 

Example Disclosure Statements

 

Editing Example 1 – Language:“This manuscript was assisted by ChatGPT (OpenAI, version 4, 2025) to improve grammar and sentence clarity. The authors reviewed and verified all suggestions to ensure factual accuracy and accept full responsibility for the content.”

:
Example 2 – Data Analysis: “A machine learning model was applied using Scikit-learn (version 1.5.0) to identify patterns within the dataset. The model was designed, validated, and interpreted by the authors, with no generative AI assistance used in drafting or analysis.”

Citing AI Tools in References

When AI tools are used for analysis, interpretation, or code generation that significantly contributed to the research, authors should cite the tool in the reference list using the following format (APA 7th Edition):

[Tool Name]. (Version). (Year). Retrieved [Date], from [URL]

Example: OpenAI. (2025, January). ChatGPT (Version 4.0) [Large language model]. https://openai.com/chatgpt

Authors should also include a brief description of how the tool was used in the methods, acknowledgements, or disclosure section of the manuscript.

If uncertain whether citation is necessary for a specific use case, authors are encouraged to consult the AJB Editorial Office prior to submission.

Failure to make such disclosure may be treated as research misconduct and can lead to immediate rejection or post-publication retraction.

 

Author Affirmation: By submitting a manuscript to AJB, all listed authors affirm that: (i) they have disclosed all AI use in accordance with this policy; (ii) they have reviewed and approved all AI-assisted content; (iii) they accept full responsibility for the accuracy, originality, and ethical integrity of the manuscript; and (iv) they understand that failure to comply with this policy constitutes research misconduct.

 

8. Editorial Oversight and Detection of AI Use

All submitted manuscripts will undergo both plagiarism and AI-content screening using specialized detection tools (e.g., Turnitin, Originality.com, and specialized AI detectors). 

AJB acknowledges that no detection tool achieves 100% accuracy; editorial judgment will be exercised in conjunction with all detection results.

Where AI use is suspected but not disclosed, editors will contact the corresponding author requesting: (i) clarification; (ii) a revised AI Use Disclosure statement; or (iii) supporting documentation (e.g., screenshots, prompts used, methodology).

If authors are found to have used AI tools in ways inconsistent with this policy, thethe Editorial Board reserves the right to :  (i) request corrections and resubmission; (ii) reject the submission; or (iii) withdraw a published article with retraction notice. In cases of serious ethical concern, the issue will be referred to AJB’s Ethics Committee for investigation in accordance with COPE flowcharts (https://publicationethics.org/). The matter may be further escalated to the author’s institutional research integrity office and relevant oversight bodies.

. 

AJB maintains a consistent process for handling suspected misconduct, with  all decisions evidence-based and documented.

 

9. Reviewer and Editor Responsibilities

Reviewers and editors share responsibility for maintaining the integrity of the peer-review process in the age of AI. Reviewers must not use AI tools to generate review reports, summaries or critiques of manuscripts unless explicitly authorized by the journal and disclosed to the editorial office. Under no circumstances should reviewers upload manuscript content to external AI systems, as this constitutes a breach of confidentiality. Editors must remain informed about evolving AI ethics and ensure that disclosure statements are assessed consistently across submissions. Where necessary, editors should guide authors in refining their disclosures to align with journal standards and ensure that AI use does not compromise the scholarly or ethical quality of the publication.

Training and Awareness: AJB is committed to ensuring consistent implementation of this policy. Editors and peer reviewers will receive periodic briefings on responsible AI use, emerging AI detection methods, and policy updates. The Editorial Office will provide training materials and guidance documents to all reviewers upon invitation. AJB aims to foster a community of practice wherein all stakeholders understand their responsibilities in maintaining research integrity in the age of AI.

 

10. Data Protection and Privacy

The African Journal of Bioethics emphasizes strict adherence to data protection laws and ethical standards. Authors and reviewers must not input human participant data, confidential peer-review materials or identifiable datasets into AI systems. All research involving human subjects must comply with national and institutional data protection frameworks, including the Kenya Data Protection Act (2019) and international standards such as the General Data Protection Regulation (GDPR). 

Global Compliance: Authors submitting from outside Kenya must comply with their respective national and regional data protection laws (e.g., GDPR for EU-based authors, Health Insurance Portability and Accountability Act [HIPAA] for US-based authors working with health data, or equivalent frameworks in their jurisdiction).

 

Authors are reminded that commercial AI models may retain data inputs for model training and improvement.ploading identifiable human participant data, confidential institutional information, or unpublished research materials to external AI systems constitutes a serious ethical and legal violation. Where AI analysis is necessary, authors should employ: (i) institutional AI systems with appropriate data governance agreements; (ii) anonymized, de-identified datasets with ethics approval; or (iii) tools with signed data processing agreements ensuring data confidentiality.

 

11. Handling of AI-Related Misconduct

Investigation Protocol:Any suspected misuse of AI tools will be investigated in accordance with the Committee on Publication Ethics (COPE) flowcharts (available at https://publicationethics.org/) and AJB’s Research Integrity Policy. 

Pre-Publication Detection: When AI-related misconduct is identified before publication, the manuscript will be desk rejected or rejected following peer review. The corresponding author will receive a written explanation of the violation and opportunity to respond within 14 days.

Post-Publication Detection: If misconduct is detected after publication, AJB will issue a retraction notice or correction notice (depending on severity) within 30 days of confirmation. The retraction will be indexed in all relevant databases.

Institutional Notification: In cases of confirmed misconduct, relevant institutional authorities (research integrity office, department head, institutional ethics committee) will be notified within 30 days of the final investigation conclusion, in accordance with COPE guidelines.

Sanctions: - First Offense: 2-year embargo on submissions; public acknowledgment of retraction. - Repeat Offenders: Permanent blacklisting from AJB submission; escalation to author’s institution and funder.

The Editorial Board will document all misconduct cases and share anonymized summaries with the Africa Bioethics Network to support sector-wide integrity standards.

 

12. Review, Updating and Implementation

This policy will be reviewed annually by the Editorial and Technical Office in consultation with the Editorial Board and external ethics advisors. Updates will reflect emerging global standards, technological advancements and evolving practices in AI-assisted research. All policy changes will be published on the AJB website and communicated to authors and reviewers through official notices.

12A. Implementation Timeline and Transition

This policy becomes effective 1 November 2025 for all new manuscript submissions via the AJB platform.

• Manuscripts submitted before 1 November 2025:
  These will be evaluated under prior guidelines. However, authors are encouraged to voluntarily comply with this policy or to provide updated AI Use Disclosure statements if their manuscripts remain under review.
  
  

• Editorial and reviewer training:
  All editors and peer reviewers will receive briefing materials and implementation guidance on this policy by 15 November 2025.
  
  

• Annual review cycle:
  This policy will be reviewed annually each October by the AJB Editorial Board, in consultation with ethics advisors and the Africa Bioethics Network, with updates published by 31 December of each review year.

 

13. Reference Frameworks and Supporting Documents

This policy draws upon and aligns with several key international frameworks, and best-practice guidelines::

• Research Ethics and Publication Standards:

• The Committee on Publication Ethics (COPE) 2024 “Guidelines on the Use of Artificial Intelligence Tools in Research and Publication.” (https://publicationethics.org/)

• The International Committee of Medical Journal Editors (ICMJE) 2023 “Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work.” (https://www.icmje.org/)

• The World Association of Medical Editors (WAME) 2023 “Recommendations on ChatGPT and Generative AI in Publication.” (https://wame.org/)

AI Ethics and Governance:

• The UNESCO (2021) “Recommendation on the Ethics of Artificial Intelligence.” (https://www.unesco.org/en/ethics-artificial-intelligence)

Publisher Policies (For Reference):

• Elsevier, Springer Nature and Taylor & Francis (2024 editions).– AI Use Policies in Manuscript Preparation and Review

Regional Data Protection:

• Kenya Data Protection Act (2019) • General Data Protection Regulation (GDPR) – EU/UK • Comparable national frameworks in authors’ jurisdictions

14. Contact Information

For questions, clarifications, or guidance regarding the use and disclosure of AI tools in manuscripts, authors are encouraged to contact the Editorial and Technical Office through the following channels:

The Editorial and Technical Office

African Journal of Bioethics (AJB)

Email: ajb@africanjournalofbioethics.org 

Website: https://www.africanjournalofbioethics.org/ 

APPENDIX: GLOSSARY OF KEY TERMS

Analytical AI: AI systems designed to analyze, interpret, or extract patterns from data (e.g., machine learning models, statistical software). Distinguished from generative AI by their primary function of data processing rather than content creation.

Authorship: The designation of human individuals who meet ICMJE criteria for substantial intellectual contribution, accountability, and approval of a scholarly work. Non-human entities (including AI systems) cannot be authors.

Disclosure Statement: A transparent, written statement documenting the name, version, purpose, and scope of AI tool use in manuscript preparation. Required in all AJB submissions where AI was employed.

Fabrication: The invention or falsification of data, results, citations, or references. A form of research misconduct explicitly prohibited in this policy.

Generative AI: AI systems trained on large datasets to generate new content, including text, images, code, or other media in response to user prompts (e.g., ChatGPT, Gemini, Claude).

Human Oversight: Active review, verification, and approval by qualified human authors of all content, including AI-generated or AI-assisted material. A core requirement of this policy.

Identifiable Data: Any information that can be used to identify an individual, including names, addresses, identification numbers, biometric data, or combinations of quasi-identifiers.

Misconduct: Any violation of this policy or established research ethics standards, including undisclosed AI use, data fabrication, or breach of confidentiality.

Plagiarism: The use of another’s words, ideas, or work without proper attribution. Using AI-generated text without acknowledgment or rewriting as original work constitutes plagiarism.

Substantial Contribution: In the context of AI use, substantial contribution is defined as: (i) >25% of any manuscript section; (ii) any portion addressing novel intellectual contribution; or (iii) content central to the work’s scholarly merit.

Transparency: Open, complete, and honest disclosure of all methods, funding, conflicts of interest, and AI use relevant to a manuscript.

 

African Journal of Bioethics – Policy on AI Use and Disclosure, Version 1, October 2025