1. Introduction

This policy has been developed in response to the increasing use of generative artificial intelligence (AI) and AI-assisted technologies in scientific and engineering research, particularly in the oil and gas sector. It provides guidance for authors, reviewers, editors, and readers of Lembaran Publikasi Minyak dan Gas Bumi (LPMGB).

LPMGB, published by the Testing Center for Oil and Gas LEMIGAS, will continue to monitor developments in this area and may revise this policy as necessary. This policy is developed with reference to internationally recognized frameworks for AI use in scholarly publishing, including the generative AI policies of Elsevier, the Committee on Publication Ethics (COPE), the International Committee of Medical Journal Editors (ICMJE) authorship principles, and guidance issued by major indexing and publishing bodies such as Scopus, Crossref, and the STM Association. As an internationally indexed journal, LPMGB holds itself to the same standard of transparency and accountability expected of leading international scientific publications.

2. Definitions and Scope

For the purposes of this policy, the term "AI Tools" refers to generative AI and AI-assisted technologies capable of producing text, code, images, data summaries, or other content in response to prompts, including but not limited to large language models (LLMs), AI writing assistants, AI-based translation and paraphrasing tools, and AI image or figure generators. This policy applies to all stages of the publishing process at LPMGB: manuscript preparation, peer review, editorial handling, production, and post-publication correction. It does not apply to conventional, non-generative software such as spelling and grammar checkers, reference managers (e.g., Mendeley, Zotero), or statistical and simulation software commonly used in petroleum engineering research, provided their use is standard and does not involve generative content creation.

3. For Authors

3.1 Use of AI Tools in Manuscript Preparation

LPMGB recognizes the potential of generative AI and AI-assisted technologies to support literature synthesis, data organization, and language improvement in manuscripts covering research, technology development, case studies, and laboratory testing in the oil and natural gas industry. Consistent with international practice, such tools may assist with improving readability and language, restructuring text for clarity, summarizing lengthy technical background material, and drafting non-critical boilerplate content (e.g., generic descriptions of standard laboratory methods). These tools must never substitute for the authors' scientific expertise, engineering judgment, and evaluative reasoning, and must not be used to generate the core scientific findings, interpretations, or conclusions of a study. Authors remain fully responsible and accountable for the content of their work, in the same manner expected by international journals such as those published by Elsevier, Springer, and Wiley.

3.2 Author Responsibilities

• Carefully verifying the accuracy, comprehensiveness, and impartiality of all AI-generated output, including technical data, formulas, units, and references, since AI-generated citations may be incorrect, outdated, or entirely fabricated ("hallucinated").
• Thoroughly editing and substantially rewriting AI-assisted material so the manuscript reflects the authors' authentic scientific analysis, interpretation, and original contribution.
• Disclosing the use of AI Tools transparently through an AI Declaration statement at submission, consistent with the disclosure requirements of Elsevier, COPE, and other major international publishers.
• Protecting data privacy, intellectual property, and confidentiality by reviewing the terms of any AI Tool used, particularly for proprietary field, reservoir, or laboratory data belonging to operators or research institutions.
• Ensuring AI-assisted translation of non-English manuscripts does not introduce factual distortion of technical terms specific to the oil and gas discipline (e.g., units, formation names, equipment terminology).
• Refraining from submitting manuscripts, or substantial portions thereof, that are predominantly AI-generated with only superficial human editing, as this may constitute a breach of research integrity.

3.3 Responsible Use of AI Tools

Authors must review the terms and conditions of any AI Tool to ensure the privacy and confidentiality of unpublished manuscripts and proprietary industry data are protected, and to confirm that the tool provider does not retain or use submitted content to train underlying models beyond what is strictly necessary to deliver the service. Particular care is required when reservoir, exploration, seismic, or laboratory data are involved, as such data may be commercially sensitive or subject to non-disclosure agreements with industry partners. AI Tools must not be used to generate images that duplicate copyrighted figures, depict real facilities or personnel, or fabricate experimental results, in accordance with the image integrity standards applied by international journals indexed in Scopus and Web of Science.

3.4 Disclosure

Authors must disclose AI Tool usage in a separate AI Declaration statement, identifying the tool name and version, the purpose of use, and the extent of human oversight applied. This declaration should be placed at the end of the manuscript, before the reference list, and may be reproduced in the published article — mirroring the disclosure placement convention adopted by Elsevier and other major international publishers. Basic grammar and spelling checks performed by standard tools do not require declaration. If AI is used within the research methodology itself (e.g., AI-assisted seismic interpretation, machine-learning-based reservoir prediction, or automated data classification), it must be described in full, reproducible detail in the Methods section, including the model name, version, training data source (where known), and validation approach.

3.5 Authorship

Consistent with ICMJE authorship criteria adopted internationally, AI Tools must not be listed as authors or co-authors and must not be cited as authors. Authorship requires accountability for accuracy, approval of the final version, agreement to submission, and the ability to respond to questions concerning the manuscript — responsibilities that can only be held by human researchers. Authors are also responsible for ensuring the work is original, not previously published, that all listed contributors meet authorship criteria, and that the manuscript does not infringe third-party intellectual property rights.

3.6 Figures, Graphs, and Laboratory Imagery

LPMGB does not permit generative AI or AI-assisted tools to create or alter figures, seismic sections, well-log plots, chromatograms, core photographs, or laboratory imagery submitted in manuscripts. This includes enhancing, obscuring, moving, removing, or introducing any specific feature within an image or figure. Adjustments to brightness, contrast, or color balance are acceptable only if they do not obscure or eliminate original data — the same standard applied by Elsevier, Nature Portfolio, and other internationally recognized publishers. LPMGB reserves the right to apply image forensics tools to detect irregularities and may request pre-adjustment or raw composite images for editorial assessment. The only exception is when AI-assisted imaging forms part of the declared research method (for example, AI-based image segmentation of core samples), in which case the tool, version, and methodology must be reported reproducibly in the Methods section. Generative AI must not be used to produce graphical abstracts or cover art without prior written permission from the editor.

4. For Reviewers

Manuscripts submitted for review must be treated as strictly confidential, in accordance with international peer review confidentiality standards. Reviewers must not upload a manuscript, any part of it, or their review report into a generative AI tool, even for the sole purpose of improving language or readability, as this may violate author confidentiality, proprietary data rights, and journal confidentiality obligations — particularly sensitive for industry-affiliated exploration and production data that may carry commercial value. Peer review is regarded internationally as a human intellectual responsibility that lies at the core of the scientific ecosystem; generative AI must therefore not be used to assist in the scientific evaluation, scoring, or recommendation of a manuscript, as it may produce incomplete, biased, or factually incorrect assessments. Reviewers remain fully responsible and accountable for the content of their review report, though they may consult the authors' AI Declaration when forming their assessment.

5. For Editors

Submitted manuscripts, editorial correspondence, notification letters, decision letters, and reviewer reports must be treated as confidential and must not be uploaded to any generative AI tool, including for language editing purposes. Editorial evaluation and decision-making require human scientific and editorial judgment and cannot be delegated to AI, consistent with the position adopted by COPE and major international publishers. Editors remain fully responsible for the editorial process, the final decision, and communication with authors. Editors may refer to the authors' AI Declaration when evaluating a submission. Suspected violations of this policy by an author, reviewer, or fellow editor should be escalated through LEMIGAS's publication ethics procedures and, where relevant, reported in line with COPE guidelines.

6. Publication Process

As part of its commitment to aligning with international publishing standards, LPMGB may explore carefully controlled uses of AI to support publication workflows, under continuous human oversight, for purposes such as:

• Identifying relevant reviewers with expertise in oil and gas research;
• Matching submissions with journal scope and section policies;
• Detecting duplicate or overlapping submissions;
• Conducting technical and completeness checks against submission requirements;
• Performing research integrity and similarity/plagiarism checks (e.g., via Turnitin);
• Supporting post-acceptance proof preparation, copyediting, and typesetting; and
• Assisting with metadata tagging and indexing for platforms such as Crossref, Scopus, DOAJ, and Google Scholar.

Human oversight remains central to decision-making throughout every stage of the publication process, and no AI-supported tool may independently determine the acceptance, rejection, or retraction of a manuscript.

7. Ethical Alignment with International Standards

This policy is consistent with broader principles of publication ethics and responsible AI use — transparency, accountability, confidentiality, human oversight, fairness, and protection of scholarly integrity — as articulated by internationally recognized bodies including:

• COPE (Committee on Publication Ethics) — guidance on authorship, transparency, and AI-related misconduct;
• Elsevier — generative AI policy for authors, reviewers, and editors, which this policy substantively references;
• ICMJE — recommendations on authorship and accountability;
• UNESCO — Recommendation on the Ethics of Artificial Intelligence;
• Nature Editorial — ground rules for the use of AI tools such as large language models in scientific writing; and
• STM Association — recommendations for classification of AI use in academic manuscript preparation.

By aligning with these frameworks, LPMGB aims to maintain the same level of trust, rigor, and transparency expected of internationally indexed journals in Scopus, Web of Science, and DOAJ.

8. Violations and Consequences

Any misuse of AI or violation of this policy may be treated as unethical conduct and handled under LPMGB's publication ethics procedures, consistent with COPE's guidelines for handling suspected misconduct. Depending on the severity of the case, actions may include manuscript rejection, mandatory revision and re-declaration, post-publication correction or retraction, notification of the authors' institutions or funding bodies, and temporary or permanent restrictions on future submissions to LPMGB.

9. Policy Review and Updates

This policy will be reviewed periodically and revised as needed to reflect developments in AI technology, evolving international standards, and emerging best practices in scholarly and technical publishing. Updated versions will be published on the official LPMGB journal website and communicated to the editorial board.

References

1. COPE (Committee on Publication Ethics). 2023. Guidance on AI and Ethics in Publishing. publicationethics.org
2. Nature Editorial. 2023. Tools such as ChatGPT threaten transparent science; here are our ground rules for their use. https://www.nature.com/articles/d41586-023-00191-1
3. UNESCO. 2021. Recommendation on the Ethics of Artificial Intelligence. https://unesdoc.unesco.org/ark:/48223/pf0000381137
4. Elsevier. 2023. AI Policy for Authors and Reviewers. https://www.elsevier.com/about/policies/artificial-intelligence
5. STM Association. 2023. Recommendations for a Classification of AI Use in Academic Manuscript Preparation. https://www.stm-assoc.org