Generative Artificial Intelligence (GenAI) Policy

The Journal of Computation Physics and Earth Science (JoCPES) recognizes the increasing use of Generative Artificial Intelligence (GenAI) and AI-assisted technologies in scientific research, manuscript preparation, programming, data analysis, and scholarly communication.

JoCPES permits the responsible use of GenAI provided that its use is transparent, ethical, verifiable, and does not replace human responsibility for the accuracy, originality, integrity, and scientific validity of the work.

Responsible Use of Generative AI

AI Tools Cannot Be Authors

Generative AI tools, large language models, chatbots, or other AI-assisted technologies cannot be listed as authors or co-authors of manuscripts submitted to JoCPES.

Authorship requires responsibility and accountability for the integrity, originality, accuracy, and final approval of the published work. These responsibilities can only be fulfilled by human authors.

Use of GenAI by Authors

Authors may use GenAI or AI-assisted technologies to support scholarly work, provided that their use complies with research-integrity and transparency requirements.

Examples of acceptable uses may include:

• language improvement and grammar correction;
• assistance with organizing or structuring text;
• programming or code-development assistance;
• data-processing or analytical assistance;
• preliminary research ideation; and
• preparation of scientific materials where appropriate.

Authors remain fully responsible for verifying all AI-assisted outputs before including them in a manuscript.

Disclosure of Generative AI Use

The use of Generative AI must be disclosed when the technology makes a substantive contribution to manuscript preparation, analysis, programming, methodology, figure generation, or other significant aspects of the research.

The disclosure should identify:

• the name of the AI tool used;
• the purpose for which it was used; and
• the extent of its contribution.

Routine spelling correction, conventional grammar checking, reference-management software, or other tools that do not generate substantive scholarly content generally do not require disclosure.

Accuracy, Originality, and Verification

Authors are fully responsible for any content produced, modified, or suggested through Generative AI.

Authors must ensure that AI-assisted content:

• is scientifically accurate;
• does not contain fabricated or misleading information;
• does not contain fabricated references or citations;
• does not plagiarize or improperly reproduce existing work;
• appropriately acknowledges original sources; and
• accurately represents the research data and findings.

References and Citations

Authors must independently verify all references suggested or generated by AI-assisted technologies.

GenAI systems may produce inaccurate, incomplete, nonexistent, or fabricated references. Fabricated citations may be treated as a serious research-integrity concern.

Scientific claims should be supported by original and verifiable scholarly sources rather than relying on AI-generated statements as authoritative references.

Data, Results, and Scientific Evidence

Generative AI must not be used to fabricate, falsify, deceptively manipulate, or misrepresent research data, experimental observations, statistical results, or scientific conclusions.

AI-generated or synthetic data may be used when scientifically justified and explicitly described as part of the research methodology. The generation procedure must be reported with sufficient transparency to support evaluation and reproducibility.

AI-Generated Images and Figures

AI-generated or AI-modified images must not be presented as genuine experimental or observational evidence when they do not represent the underlying scientific data.

When GenAI is used to generate or substantially modify a scientific illustration, figure, graphical element, or visualization, authors must disclose the use of the technology and ensure that the resulting material does not misrepresent the research.

AI-Assisted Programming and Scientific Computing

Because JoCPES publishes research involving scientific computing, numerical methods, machine learning, and related computational approaches, AI-assisted programming tools may be used responsibly.

Authors remain responsible for:

• verifying the correctness of generated code;
• validating computational results;
• documenting important algorithms and methods;
• ensuring appropriate software licensing;
• identifying potential methodological or security issues; and
• providing sufficient information to support reproducibility.

When AI-generated code constitutes a substantial component of the research methodology, its use should be disclosed in the manuscript.

Use of Generative AI by Reviewers

Submitted manuscripts are confidential documents. Reviewers must not upload, paste, transmit, or otherwise provide unpublished manuscripts, figures, data, supplementary files, or confidential editorial information to publicly accessible AI systems when doing so may compromise confidentiality, intellectual property, or author privacy.

Reviewers remain fully responsible for the accuracy, fairness, and integrity of their review reports. GenAI must not replace independent scientific judgment.

Substantive use of AI-assisted technologies in preparing a review should be disclosed to the handling editor where appropriate.

Use of Generative AI by Editors

Editors must protect the confidentiality of submitted manuscripts and should not provide unpublished manuscript content to publicly accessible AI systems where confidentiality cannot be assured.

AI-assisted technologies may support administrative or language-related editorial activities, but final editorial decisions must always be made by qualified human editors.

AI-generated assessments must not independently determine whether a manuscript is accepted, revised, or rejected.

AI Detection Tools

JoCPES may use technological tools to assist in identifying potentially AI-generated, manipulated, plagiarized, or otherwise problematic content.

Automated AI-detection results will not normally be used as the sole basis for an editorial decision. Concerns identified by automated tools should be considered together with editorial assessment and other available evidence.

Violations of This Policy

Failure to disclose significant GenAI use, fabrication of references or data, deceptive AI-generated content, confidentiality breaches, or other inappropriate uses of AI may be investigated as potential publication-ethics violations.

Depending on the circumstances, JoCPES may:

• request clarification or correction;
• require manuscript revision;
• reject the manuscript;
• correct a published article;
• issue an Expression of Concern;
• retract a published article; or
• take other appropriate editorial action.

Policy Updates

Generative Artificial Intelligence technologies and scholarly publishing standards continue to evolve. JoCPES may periodically update this policy to reflect developments in research integrity, publication ethics, and international scholarly publishing practices.