Generative AI and AI-Assisted Technologies Policy

Hibiscus Publisher Generative AI and AI-Assisted Technologies Policy

1. Objective and Applicability

As generative artificial intelligence (GenAI) and related AI-driven tools become embedded in academic research and scholarly communication, they offer valuable support for tasks ranging from data analysis and coding to translation and manuscript drafting. However, the integration of these tools introduces critical ethical challenges regarding intellectual property, privacy, research integrity, and accountability.

Hibiscus Publisher acknowledges the utility of these technologies while firmly upholding that transparent reporting, scholarly judgment, and human accountability are irreplaceable cornerstones of the academic publication process. This document outlines the mandatory standards for utilizing GenAI tools across all Hibiscus Publisher publications. It applies universally to authors, peer reviewers, editors, and all other participants in the publication workflow. Individual journals retain the right to enforce supplementary guidelines tailored to their specific disciplines, provided they align with this foundational policy.

2. Terminology

  • Generative AI (GenAI): Systems engineered to autonomously produce or heavily modify diverse content formats—including text, audio, video, computer code, and images—based on user prompts (e.g., large language models, chatbots, and image generators).
  • AI-Assisted Technologies: Software leveraging artificial intelligence to facilitate specific academic tasks, such as language refinement, summarization, literature retrieval, or image processing.
  • AI as a Research Mechanism: Utilizing AI as an experimental focus, analytical methodology, computational instrument, or object of study is fundamentally distinct from using it to draft a manuscript. This policy does not prohibit the methodological use of AI in research, provided it is comprehensively detailed in the manuscript.

3. Guiding Principles

  • Absolute Human Accountability: The human user bears ultimate responsibility for all published material, editorial decisions, and research outcomes. Liability and scholarly responsibility cannot be transferred to an AI system.
  • Mandatory Transparency: Any application of GenAI or AI-assisted technologies must be explicitly disclosed according to these guidelines.
  • Rigorous Verification: Because AI models are prone to generating biased, fabricated, or hallucinated information (including fake citations), users are required to independently validate all AI-generated outputs.
  • Safeguarding Integrity: AI must never be leveraged to forge data, falsify evidence, fabricate references, or facilitate publication misconduct.
  • Data Privacy and Confidentiality: Users are strictly forbidden from inputting unpublished manuscripts, sensitive personal data, or confidential peer-review materials into external AI platforms.
  • Primacy of Human Judgment: While AI can streamline administrative and editorial workflows, it cannot substitute for the critical evaluation and intellectual rigor of a human scholar.
  • Denial of AI Authorship: AI systems do not possess legal or moral agency and cannot be credited as authors.

4. Author Guidelines for Generative AI

4.1 Ethical Utilization

Authors are permitted to utilize GenAI tools for legitimate preparatory tasks, provided such use does not compromise intellectual property, privacy, or research integrity. Acceptable applications include streamlining grammar, enhancing readability, translating text, structuring arguments, or assisting with code and data analysis. Authors must never allow AI to replace their own analytical thinking, interpretation, or intellectual contribution. The author assumes total responsibility for editing, verifying, and correcting any AI-assisted output.

4.2 Mandatory Declaration Requirement

To ensure an accurate article-level record of AI utilization, every manuscript submitted to a Hibiscus Publisher journal must include a formal declaration, regardless of whether AI tools were employed.

  • When AI Was Not Used: Authors must include the following standard statement:

Declaration of Generative AI and AI-Assisted Technologies: The authors declare that no Generative AI or AI-assisted technologies were used in the preparation of this manuscript or in the research process reported in this article.

  • When AI Was Used: Authors must explicitly name the technology and detail its application using the following template:

Declaration of Generative AI and AI-Assisted Technologies: During the preparation of this manuscript, the authors used [Name of Tool] for [Description of Use]. The authors reviewed and verified the AI-assisted output, made the necessary revisions, and take full responsibility for the accuracy, originality, integrity, and final content of the manuscript.

4.3 Disclosure Thresholds

Explicit disclosure is required if AI was used to significantly generate, translate, reorganize, or rewrite manuscript text, figures, tables, reference lists, or computer code. Routine grammar and spell-checking software that does not materially alter the scholarly substance does not require specific tool identification, though the baseline article-level declaration remains mandatory.

4.4 AI within the Research Methodology

If AI constitutes a core component of the experimental design, data analysis, or software development, this must be thoroughly documented in the Methods section (or equivalent). The disclosure should include the model version, developer, date of access, and precise operational parameters to allow for peer evaluation.

4.5 Visual Media and AI Generation

Authors must guarantee that any AI-modified or AI-generated visual content does not misrepresent findings, fabricate results, or violate copyright laws. If AI image processing materially impacts the presentation of evidence, it must be comprehensively detailed. Hibiscus Publisher journals may impose stricter limitations on AI-generated artwork depending on disciplinary standards.

4.6 Citations and Bibliographies

AI systems frequently hallucinate nonexistent identifiers (DOIs), fabricated authors, and phantom publications. Authors must independently verify every cited source against the original publication. AI output cannot be used as proof that a specific claim is supported or that a source exists.

5. Authorship Parameters

To qualify for authorship, an entity must be capable of approving final drafts, responding to scholarly critiques, and taking legal/ethical responsibility for the work. Consequently, GenAI systems cannot hold authorship status. While their assistance must be declared, this does not equate to authorship.

6. Peer Reviewer Obligations

6.1 Strict Confidentiality

Peer review operates on strict confidentiality. Reviewers are expressly prohibited from uploading submitted manuscripts, supplementary data, or editorial correspondence into any external GenAI system (even those marketed as "private" or "secure").

6.2 Assessment and Reporting

Reviewers must not outsource their substantive scholarly assessment to an AI program. An AI cannot dictate manuscript acceptance, methodological validity, or novelty. Reviewers may utilize approved AI tools solely to refine the grammar or structure of their own written feedback, provided no confidential manuscript data is inputted and the core judgment remains entirely human.

7. Editorial Workflow and Responsibilities

7.1 Maintaining Confidentiality

Editors are bound by the same strict confidentiality protocols as reviewers and must not transfer unpublished manuscript materials to unauthorized external AI platforms.

7.2 Decision-Making Authority

While editors may leverage AI to assist with administrative functions, the ultimate decision regarding a manuscript's acceptance, rejection, or ethical standing must always originate from a qualified human editor.

7.3 Automated Screening Procedures

Hibiscus Publisher reserves the right to employ specialized AI tools for metadata verification, plagiarism screening, image manipulation detection, and technical compliance. However, AI-generated flags or similarity scores are strictly advisory; a human editor must interpret these metrics before taking any consequential punitive or editorial action.

8. Misconduct and AI Detection

Violations of these guidelines—such as concealing AI usage, fabricating data via GenAI, uploading confidential manuscripts to external models, or falsely attributing AI-generated text as original work—constitute severe academic misconduct. Hibiscus Publisher may utilize AI-detection software during investigations; however, a high detection score will not serve as standalone conclusive evidence of misconduct without broader contextual evaluation by the editorial board.

9. Policy Evolution

Given the rapid advancement of artificial intelligence technologies, Hibiscus Publisher will continuously evaluate and update this framework to ensure alignment with international standards set by bodies such as COPE, WAME, and STM. The active iteration published on the official Hibiscus Publisher portal acts as the binding policy document.

10. Industry Standards & Standardized Disclosure Practices

When formulating disclosures, authors should adhere to established norms enforced by leading scientific publishers and citation styles regarding generative AI usage:

10.1 Text and Prompt Disclosures

Across major publishing houses, routine grammar corrections generally do not require specific AI model declarations. However, substantive drafting requires explicit prompt transparency.

  • APA Style Guidance: APA treats the company that developed the AI (e.g., OpenAI, Google) as the author. The reference should include the specific date the chat occurred, a description of the generated content (often drawn from the prompt), the specific model version (e.g., GPT-4o), and a stable, shareable chat link if the tool permits it.
    • Example Prompt Disclosure Statement: "ChatGPT-4 (OpenAI, version August 2026) was utilized to summarize qualitative interview themes. The following prompt was used: 'Summarize the primary recurring themes in the following transcript excerpts regarding...' The authors subsequently verified all extracted themes against the original raw data."
  • IEEE Style Guidance: IEEE does not utilize a formal reference list entry for AI-generated text. Instead, authors must disclose the use of the AI system within the acknowledgments section, explicitly naming the tool, the specific sections of the paper it contributed to, and a brief explanation of how it was utilized.

10.2 Policies on AI-Generated Figures and Tables

The use of AI to generate visual data representations (infographics, plots, tables, and images) is strictly regulated and varies among publishers. Regardless of the publisher, all AI-generated images should include the AI tool name, the specific prompt used, the generation date, and a retrieval link.


Authors must explicitly declare if any AI tool was used in the creation of images or figures. While AI-assisted data visualization (e.g., using AI to format a chart based on real data) is generally permitted with disclosure, generating conceptual illustrations receives high scrutiny, and AI-generated content cannot be listed as an author. All AI-generated visual content must be declared.

It is strictly prohibited to use AI to generate or alter primary experimental data imagery (such as microscopy, gels, or blots). AI-generated images are generally restricted to schematic diagrams or graphical abstracts, which must be clearly labeled as AI-generated in the image caption. Elsevier requires disclosure of these tools in both the image caption and the general AI disclosure statement within the manuscript.

Examples

Declaration of Generative AI and AI-Assisted Technologies

During the preparation of this manuscript, the authors utilized the following AI-assisted technologies:

  • Language and Editing Assistance: The authors used Grammarly (Grammarly Inc., version 7.0) to review and refine the manuscript for grammar, punctuation, and structural clarity.
  • Drafting and Conceptualization: ChatGPT-4 (OpenAI, version August 2026) was used to assist in brainstorming the initial structural outline and summarizing key themes in the literature review section.
  • Data Extraction and Synthesis: Claude Sonnet 3.5 (Anthropic, version April 2026) was utilized to systematically extract specific qualitative themes and achievements from the selected literature corpus to aid in data synthesis.

Verification Statement:

All AI-generated outputs, summaries, and extracted data were independently verified by the authors against the original raw data and source materials. The authors reviewed, substantially edited, and revised all AI-assisted text. No AI tool was used to generate new data, alter original research findings, or create primary visual data representations. The authors take full responsibility for the accuracy, originality, and final scholarly content of this manuscript.