The rapid evolution of Generative Artificial Intelligence (GenAI), including Large Language Models (LLMs), is fundamentally reshaping the landscape of scientific communication. While these technologies are recognized for their potential to augment researcher productivity—particularly in conceptual brainstorming and linguistic refinement for non-native English speakers—their application must be governed by the principles of accountability, transparency, and scholarly integrity.
1. Authorship and Accountability
In accordance with established scholarly standards, AI-based tools cannot be credited as authors or co-authors. Authorship implies legal and ethical responsibilities that AI entities cannot fulfill.
Ultimate Responsibility
Authors remain exclusively accountable for the entire content of their manuscript, including the accuracy of technical claims, methodological soundness, and adherence to ethical standards.
Integrity
Authors are responsible for preventing plagiarism and ensuring that AI-assisted content does not infringe upon existing copyrights or intellectual property.
2. Mandatory Disclosure and Transparency
Transparency is a prerequisite for the ethical use of GenAI. Authors must explicitly disclose any AI involvement within the manuscript (typically in the acknowledgments or a dedicated “Disclosure of AI Usage” section).
Minimum expectation
Clearly state what tools were used and for what purpose (e.g., grammar polishing, paraphrasing under supervision), without implying AI as an author.
3. Data Integrity and Prohibition of Misconduct
The use of GenAI to fabricate or falsify research data, experimental results, source code, or scientific visualizations is strictly prohibited. Such actions are classified as research misconduct and may lead to the immediate rejection of the manuscript or post-publication retraction.
Verification
All AI-generated outputs—particularly those used in literature reviews—must be manually verified against primary sources to eliminate hallucinations or factual inaccuracies.
4. AI in Figures, Illustrations, and Image Processing
The application of GenAI to visual media is permitted only under stringent conditions:
Illustrative Use
AI-generated images (e.g., conceptual diagrams) must be explicitly labeled as “Illustrative” in the figure caption to prevent them from being mistaken for empirical data.
Image Modification
AI-assisted editing (e.g., removing background noise) is acceptable provided it does not alter the scientific interpretation of the data. All such modifications must be disclosed in the caption.
Legal Compliance
Authors must ensure that AI-generated visual content does not violate copyright laws.
5. Linguistic Refinement and Formatting
AI tools may be ethically employed for stylistic and linguistic improvements, such as correcting grammar, syntax, and orthography; condensing text to meet specific word-count constraints; and translating preliminary drafts into scholarly English. Despite these enhancements, the final prose remains the intellectual product of the authors, who must ensure that the nuanced meaning of the research is preserved.
