AI Writing
Artificial Intelligence (AI) Policy
The Journal of Agriculture and Food Science Applications (JAFSA) recognizes the legitimate role of artificial intelligence and AI-assisted technologies in research, scholarly communication, editing, data analysis and publishing. At the same time, JAFSA requires responsible, transparent and accountable use of such technologies to protect research integrity, originality, confidentiality and public trust.
This policy applies to authors, reviewers, editors, editorial staff, guest editors, and other contributors involved in the submission, peer-review, editing, production and publication of manuscripts in JAFSA. It applies to generative AI, large language models (LLMs), chatbots, AI writing assistants, AI image generators, AI coding tools, AI data-analysis systems and other AI-assisted technologies.
JAFSA's policy is based on the principles of human accountability, transparency, research integrity, confidentiality, appropriate attribution, protection of intellectual property and responsible scientific communication.
1. Core Principles
Human Accountability
AI tools cannot assume responsibility for a manuscript. Human authors remain fully responsible for the accuracy, originality, integrity, interpretation and conclusions of their work.
Transparency
Authors, reviewers and editors must disclose substantive use of AI when such use affects the content, analysis, figures, images, code, interpretation or evaluation of scholarly work.
Research Integrity
AI must not be used to fabricate, falsify, manipulate, conceal or misrepresent research data, references, images, results or conclusions.
2. AI Tools Cannot Be Authors
JAFSA does not recognize artificial intelligence systems, chatbots or other AI tools as authors or co-authors of scholarly articles. Authorship carries responsibility for the work, including accountability for its accuracy, integrity and originality. AI systems cannot accept such responsibility.
Authors must therefore ensure that every person listed as an author meets the journal's authorship requirements and has approved the final manuscript. The use of an AI tool does not constitute authorship.
3. Acceptable Use of AI by Authors
JAFSA does not prohibit responsible use of AI-assisted technologies. Authors may use AI tools where the use is appropriate, transparent and consistent with research integrity.
Examples of potentially acceptable uses include:
- Grammar, spelling, punctuation and language improvement.
- Improving readability and clarity of author-written text.
- Translation or language assistance, provided the translated content is carefully verified.
- Generating programming or coding suggestions that are subsequently checked by the authors.
- Assistance with data analysis where the methodology, data and outputs are independently verified.
- Generating ideas, outlines or possible research questions, provided the final intellectual contribution is made by the authors.
- Assistance with formatting, organization or non-substantive editorial tasks.
- Generating preliminary visual concepts, provided that the resulting figures or images are scientifically accurate and appropriately disclosed.
4. AI-Generated or AI-Assisted Text
Authors may use AI-assisted writing tools for limited language and editorial assistance. However, authors must carefully examine all AI-generated material before submission.
Authors are responsible for identifying and correcting hallucinated information, inaccurate statements, unsupported claims, fabricated references, biased interpretations, inappropriate terminology and other errors generated by AI.
AI-generated text must not be presented as though it represents independent human scholarship when it materially contributes to the intellectual or substantive content of the manuscript.
5. AI Use in Research, Data Analysis and Interpretation
When AI is used to collect, process, classify, analyze, model or interpret research data, the use must be disclosed sufficiently to allow readers, reviewers and editors to understand its role in the research.
Where relevant, authors should report:
- The name of the AI system or software.
- The developer or provider, where relevant.
- The model or version used.
- The date or period of use.
- The purpose for which the system was used.
- The nature and source of the input data.
- Important prompts, instructions or configurations where necessary for reproducibility.
- Human verification, validation or quality-control procedures.
- Any material limitations, uncertainty or potential bias introduced by the AI system.
AI-generated analytical results must not be accepted without appropriate scientific validation. Authors remain responsible for determining whether AI-assisted analyses are statistically, scientifically and methodologically appropriate.
6. AI-Generated Figures, Images and Illustrations
AI-generated or AI-enhanced images, illustrations, diagrams and graphical materials must be used responsibly and must not misrepresent experimental evidence.
Prohibited practices include:
- Creating fictional experimental results and presenting them as real observations.
- Adding, removing or altering research features in a manner that changes the scientific interpretation.
- Manipulating microscopy, field, laboratory, agricultural or experimental images to create misleading evidence.
- Generating images that are falsely presented as photographs of actual experiments, specimens, field sites or equipment.
- Using copyrighted images through AI without appropriate rights or attribution.
If an AI tool is used to generate or materially modify a figure or image, authors should disclose the tool, version and nature of the modification in the manuscript or figure documentation, where appropriate.
7. AI-Generated References and Citations
Authors must independently verify every reference included in a manuscript. AI systems can generate references that are inaccurate, incomplete, incorrectly attributed or entirely nonexistent.
Fabricated, unverifiable or incorrectly attributed references are unacceptable. Authors must consult the original source and ensure that the cited publication actually supports the claim for which it is cited.
AI-generated content must not be treated as a substitute for appropriate primary scientific literature. Where AI-assisted tools are used to identify literature, authors remain responsible for conducting appropriate literature verification and citation.
8. AI and Plagiarism
The use of AI does not exempt authors from JAFSA's plagiarism, duplicate publication, copyright or research-integrity requirements.
Authors must ensure that AI-assisted text, images, code, tables and other material do not reproduce protected or previously published material without appropriate permission, attribution and citation.
Authors remain responsible for any plagiarism, unattributed reproduction, fabricated references, copyright infringement or other ethical violation contained in AI-assisted material.
9. Required AI Disclosure by Authors
Authors must disclose substantive use of AI-assisted technologies during manuscript preparation or research. The disclosure should be sufficiently specific to enable editors and readers to understand what role AI played.
Recommended Disclosure Statement
Where AI was used only for basic spelling, grammar or punctuation correction, authors should nevertheless follow the journal's submission disclosure requirements and provide disclosure whenever the use is substantive or requested by the editorial office.
10. AI Disclosure by Authors: What Must Be Reported?
| AI Use | JAFSA Position | Disclosure |
|---|---|---|
| Basic spelling and grammar correction | Generally acceptable | Disclose when substantive or requested |
| Language polishing or translation | Acceptable with human verification | Yes |
| Generating substantive manuscript text | May be permitted with transparency | Yes |
| Generating or transforming research data | Requires rigorous justification and validation | Yes |
| AI-assisted statistical/data analysis | Acceptable when scientifically validated | Yes |
| AI-generated figures or illustrations | Subject to scientific and copyright review | Yes |
| Fabricating research results with AI | Prohibited | Not applicable |
| Listing AI as an author | Prohibited | Not applicable |
| Generating fake references | Prohibited | Not applicable |
11. Confidentiality and Sensitive Information
Authors, reviewers, editors and editorial staff must exercise particular caution before entering confidential, unpublished or personally identifiable information into external AI systems.
This requirement is particularly important where manuscripts contain unpublished research, proprietary information, personal data, commercially sensitive information, intellectual property or information belonging to research participants or third parties.
12. Artificial Intelligence and Peer Review
JAFSA recognizes that AI tools may assist reviewers with certain administrative or language tasks. However, reviewers must protect the confidentiality of manuscripts and must not use AI in a manner that exposes confidential manuscript content.
Reviewers must not:
- Upload a confidential manuscript to a public or external AI system without authorization from JAFSA and adequate confidentiality safeguards.
- Use AI to generate a review without appropriate disclosure and human evaluation.
- Allow AI-generated recommendations to replace their independent scholarly judgment.
- Use confidential manuscript information to train, prompt or develop an AI system.
- Use AI to appropriate ideas, methods, data or findings contained in an unpublished manuscript.
A reviewer who uses an AI tool in connection with a review must disclose that use to the editorial office and remains fully responsible for the accuracy, fairness and confidentiality of the review.
13. Artificial Intelligence and Editors
Editors may use appropriate AI-assisted technologies for limited administrative, language, workflow or screening purposes where such use does not compromise confidentiality, editorial independence or research integrity.
Editorial Responsibility
Editors remain responsible for editorial decisions. AI recommendations must never automatically determine acceptance, rejection, reviewer selection or other substantive editorial outcomes without appropriate human oversight.
Confidentiality
Editors and editorial staff must not expose confidential manuscripts or reviewer materials to AI systems unless confidentiality and data protection can be adequately assured.
Accountability
Human editors remain accountable for editorial correspondence, decisions and published material, including content produced with AI assistance.
Independence
AI tools must not replace qualified editorial judgment or introduce discriminatory, biased or unexplained criteria into editorial decision-making.
14. AI-Assisted Editorial Screening and Detection
JAFSA may use appropriate technological tools, including AI-assisted systems, similarity checking tools or other screening technologies, to support the editorial assessment of submitted manuscripts.
Such tools are considered screening aids rather than definitive evidence. An AI-generated-content detector or automated screening result alone will not necessarily establish misconduct.
Where concerns arise, the editorial office may conduct additional human assessment, request clarification or supporting information from authors, examine the underlying research and references, or take other appropriate editorial action.
15. AI and Research Integrity Violations
The following uses of AI may constitute research or publication misconduct when undertaken deceptively or in violation of applicable ethical standards:
- Fabricating or falsifying research data.
- Inventing experimental results or observations.
- Generating false or nonexistent references.
- Manipulating images to misrepresent scientific findings.
- Creating fictitious participants, samples, experiments or field observations.
- Concealing substantive AI assistance where disclosure is required.
- Using AI to plagiarize or reproduce protected material without appropriate attribution.
- Misrepresenting AI-generated content as independently conducted research.
- Using AI to circumvent peer review, research-integrity checks or editorial safeguards.
- Submitting AI-generated material containing fabricated claims or unsupported conclusions.
16. Consequences of Non-Disclosure or Misuse
Where JAFSA becomes aware of undisclosed or inappropriate AI use, the editorial office may request an explanation and supporting information from the authors or other parties involved.
Depending on the circumstances and seriousness of the matter, possible actions may include:
- Requesting correction or clarification before publication.
- Requiring revision of the manuscript or disclosure statement.
- Rejecting the manuscript.
- Suspending or terminating the peer-review process.
- Issuing a correction, clarification or editorial notice after publication.
- Retracting or withdrawing an article where warranted.
- Notifying the relevant institution or other appropriate body where serious misconduct is suspected.
- Taking other actions consistent with JAFSA's publication-ethics procedures.
The response will depend on the nature, extent and impact of the suspected violation and will be handled in accordance with applicable publication ethics and due-process principles.
17. Responsibility for AI-Assisted Content
The use of AI does not transfer responsibility from the authors to the software developer, AI provider, publisher or journal.
The submitting and corresponding authors are responsible for ensuring that AI-assisted content is accurate, original, appropriately attributed, ethically produced and consistent with the underlying research.
18. Protection of Personal and Research Data
Authors must comply with applicable data-protection, privacy, research-ethics and confidentiality requirements when using AI systems in research. Personal, confidential or sensitive research data should not be entered into external AI systems unless there is a lawful and ethically appropriate basis for doing so and adequate safeguards are in place.
Where research involves human participants, authors remain responsible for complying with applicable ethical approval, informed-consent and confidentiality requirements. AI use must not undermine commitments made to research participants.
19. Generative AI and the Integrity of Agricultural Research
Because JAFSA publishes research relating to agriculture, food science, food safety, biotechnology, environmental management, livestock, fisheries, aquaculture, agricultural engineering and related disciplines, authors should exercise particular caution when AI is used to generate scientific recommendations or practical conclusions.
AI-generated recommendations concerning agricultural production, animal health, food safety, chemical use, biotechnology, environmental management, nutrition or other potentially consequential areas must be independently verified against appropriate scientific evidence.
Authors should not present an AI system's output as scientific evidence merely because the output appears authoritative or technically plausible.
20. Editorial Review of AI Disclosures
JAFSA editors may request additional information where an author's AI disclosure is unclear or where the editorial team reasonably believes that AI may have played a substantive role in the manuscript.
Authors should retain sufficient information about substantive AI use to explain how the technology contributed to the work. Where appropriate, this may include the tool name, version, dates of use, prompts, relevant outputs, validation procedures or other documentation.
21. Recommended Author Checklist
Before submitting a manuscript to JAFSA, authors should confirm that:
- All authors remain responsible for the submitted work.
- No AI system has been listed as an author or co-author.
- All substantive AI use has been appropriately disclosed.
- AI-generated claims have been independently checked.
- All references have been verified against the original sources.
- No fabricated citations or sources have been included.
- AI-assisted text does not contain unattributed plagiarism.
- AI-generated or AI-enhanced figures accurately represent the underlying research.
- AI has not been used to fabricate or falsify research data.
- Confidential or sensitive information has not been improperly submitted to an external AI system.
- The manuscript accurately describes AI-assisted methods where AI materially contributed to research or analysis.
- The authors can explain and defend all substantive aspects of the submitted work.
22. Ongoing Review of This Policy
AI technologies and their applications in scholarly communication are evolving rapidly. JAFSA may periodically review and update this policy to reflect developments in research ethics, publishing standards, data protection, copyright, AI technology and responsible scholarly communication.
Any updated policy will apply to submissions and publication activities according to the effective date specified by the journal.
23. Relevant Publishing Guidance
JAFSA's AI policy is informed by established principles concerning transparency, human accountability, authorship, confidentiality and responsible use of AI in scholarly publishing.