Policy on AI and LLMs

Policy on Artificial Intelligence (AI) and Large Language Models (LLMs)

1. Purpose and Scope

This policy establishes the governing standards for the integration of Artificial Intelligence (AI) and Large Language Models (LLMs) within the ETHIOINQUIRY publishing framework. Its primary objective is to safeguard the integrity, transparency, and trust of scholarly records. This policy applies to all authors, peer reviewers, and editorial board members of the journal. While ETHIOINQUIRY acknowledges that technology evolves, the fundamental responsibilities of scholarship remain human-centric; technological assistance does not diminish the personal accountability of human researchers involved.

2. Authorship and Attribution Criteria

Large Language Models (LLMs) and generative AI tools do not satisfy the criteria for authorship. Consequently, they shall not be listed as authors or co-authors of any manuscript submitted to ETHIOINQUIRY.

Reasons for AI Authorship Prohibition

  • Accountability: Authorship implies legal and ethical responsibility for the research conducted. AI tools cannot be held accountable for the accuracy, integrity, or originality of their work.
  • Legal Consent and Agreements: AI tools lack the legal standing to provide consent, sign Conflict of Interest (COI) declarations, or enter into binding copyright agreements.
  • Professional Integrity: Authorship provides credit for substantial intellectual contributions, a standard reserved exclusively for human researchers.
  • Verification of Findings: Only human authors can testify to the veracity of the data and ensure that the final version reflects their original intentions.

Citations and Attribution: Authors are strictly forbidden from citing AI tools as authors in the bylines or formal reference lists. However, if an AI tool or non-generative machine learning software is used as a resource for data analysis or processing, it should be cited as a software resource in the appropriate section, distinct from the authorship.

3. Guidelines for Authors: Usage and Disclosure

ETHIOINQUIRY mandates a strict distinction between AI-assisted technical refinement and restricted, generative content creation.

AI Usage Framework

Acceptable AI-Assisted Copy-Editing Restricted Generative Content Creation
Improving the readability, style, and flow of human-generated texts. Autonomous generation of core manuscript content, hypotheses or findings.
Correcting grammar, spelling, punctuation, and linguistic tone. Editorial work that changes the substantive meaning or interpretation of an original work.
Refining the formatting of the original human-authored text. Producing new interpretations or interpretations of data without human origination.

Disclosure Requirements

If AI tools are utilized in the preparation of a manuscript beyond basic copy editing, authors must adhere to the following:

  1. Documentation Location: Report the use of AI in the Methods section. If a Methods section is not available, disclosure must be provided in a suitable alternative part of the manuscript, such as the Acknowledgments section.
  2. Required Specification: The disclosure must explicitly state the tool name, version, and specific nature of the assistance provided.
  3. Human Verification: Authors must declare that the AI-assisted edits accurately reflect their original work and that they are in full agreement that the final version reflects their original intent.

Generative AI Images and Media

ETHIOINQUIRY prohibits the use of generative AI-created images and videos because of unresolved copyright and research integrity concerns.

  • Exceptions: Exceptions are granted only for the following:
  1. Images obtained from agencies with which the publisher has a contractual relationship, provided that the images were created in a legally acceptable manner.
  2. Images/videos that are the primary subject of a piece specifically discussing AI technology.
  3. Generative tools trained on verified and attributable scientific data.
  • Disclosure of Non-Generative ML: The use of non-generative machine learning tools to manipulate, combine, or enhance existing images or figures must be disclosed in the relevant captions to allow for a case-by-case review.
  • Labeling: All AI-influenced media must be clearly labeled as "AI-generated" or "AI-enhanced" within the image field or the caption.

4. Human Accountability and Responsibility

Authors are 100% accountable for the accuracy and integrity of the final manuscript.

  • Verification Mandate: Authors must manually verify all AI-assisted output to ensure that the manuscript is free of "hallucinations," biased information, or false data often produced by generative tools.
  • Originality: Authors must ensure that any AI-assisted refinements do not infringe upon the intellectual property of others or breach plagiarism standards.

5. Ethical Mandates for Reviewers and Editors

To preserve the confidentiality of the peer-review process, ETHIOINQUIRY enforces the following mandates.

Confidentiality Protections

Reviewers and editors are strictly prohibited from uploading submitted manuscripts or any part thereof into generative AI systems.

Warning: The expertise provided by the peer reviewers is invaluable and irreplaceable. Reviewers are selected for their in-depth knowledge, which AI cannot replicate. Generative AI tools lack up-to-date knowledge and are prone to producing biased, incorrect, or nonsensical information. Furthermore, manuscripts contain sensitive or proprietary information; uploading them to external AI tools compromises the confidential peer-review process, as data security and privacy cannot be guaranteed.

Editorial Transparency

  • Accessory Content: Editors may utilize approved AI tools to support the generation of accessory content such as social media posts, plain language summaries, editorial summaries, and glossary terms. All such content must be fact-checked and edited by human staff members.
  • Human Decision-Making: While AI may assist in administrative tasks such as metadata retrieval or summarization, the final decision-making power regarding the evaluation and acceptance of manuscripts rests solely with human editors.

6. Compliance, Corrections, and Research Integrity

Violations of this policy, including the non-disclosure of AI use, constitute a breach of research integrity and will result in formal action.

Consequences of Non-Disclosure and Misconduct

  1. Rejection: For manuscripts under consideration in which policy breaches are identified.
  2. Expression of Concern: Issued for published works if an investigation is underway or if the evidence is inconclusive but indicates potentially unreliable findings or serious ethical concerns.
  3. Retraction: Issued in cases of major error (e.g., miscalculation), fraudulent use of data, falsification (including AI-driven image manipulation), or when the editor has lost confidence in the results due to undisclosed AI use.

The Scholarly Record

ETHIOINQUIRY maintains a permanent and unaltered historical record of scholarship. To ensure total transparency, any correction involving figures or tables will be issued via an amendment notice that reproduces the original data alongside the corrected version, allowing the scientific community to track the precise nature of the changes.

7. Review and Evolution Statement

ETHIOINQUIRY continuously monitors rapid developments in AI technology. This policy will be reviewed and updated regularly to remain aligned with the evolving best practices and industry standards established by the Committee on Publication Ethics (COPE) and the International Committee of Medical Journal Editors (ICMJE).

⬇ Download AI & LLM Policy