|
Guidelines for Authors Regarding the Use of Artificial Intelligence Tools
AI Does Not Qualify for Authorship
No Artificial Intelligence (AI) tools, such as ChatGPT, Claude, Midjourney, or similar Large Language Models (LLMs), can be credited as an author or co-author of a manuscript. This is because AI lacks the capacity to assume the legal and ethical responsibilities required for authorship.
Transparency and Disclosure of Use
If authors utilize generative AI for translation, language editing, or scientific data extraction, they must transparently disclose this usage in the manuscript. This disclosure is typically placed at the end of the article (prior to the References section) under a heading such as "Declaration of Generative AI in Scientific Writing," or within the "Methodology" section. The declaration must specify the name of the tool, its version, and the exact nature of its application.
Authors Hold Full Responsibility for Content
Human authors assume 100% responsibility for the accuracy, integrity, originality (absence of plagiarism), and objectivity (lack of bias) of the entire manuscript. AI "hallucinations" and fabricated citations constitute research misconduct. If an AI tool extracts erroneous information from a source (due to hallucination) and it is incorporated into the manuscript, the author(s) are held entirely accountable.
Generation of Images, Data, and Content
The generation of images, graphs, raw data, and structural written content by AI is strictly prohibited.
Unacceptable Uses of Generative AI (Research Misconduct)
Manuscript Writing
- Drafting fundamental sections of the manuscript (e.g., Abstract, Introduction, Literature Review, Methodology, and Conclusion) using AI is strictly prohibited.
- Copy-pasting text generated by chatbots into any section of the research is forbidden.
- Utilizing AI tools to bypass or obfuscate plagiarism detection software is prohibited.
Data Generation
- Under no circumstances should AI be the generator of primary raw data.
| Fabrication or manipulation of raw data is strictly prohibited. This includes fabricating questionnaire and statistical data, environmental, field, and laboratory data, qualitative data (e.g., fake interviews or fabricated documentary evidence), and visual data (unrealistic generation, manipulation, or alteration of images, maps, field photographs, and charts by AI). |
Content Generation
- Generating the final research idea, hypothesis formulation, scientific inference, and interpretation of results are strictly prohibited. Machine outputs must not substitute the critical thinking of the researcher.
|
Ideas and hypotheses must stem from a genuine research gap identified through a comprehensive literature review or active fieldwork. Machine-generated hypotheses merely represent statistical repetitions of historical text patterns and lack genuine academic novelty.
The use of AI in interpreting results, scientific deduction, and drawing final conclusions is strictly prohibited. All qualitative and quantitative analyses, as well as scientific reasoning, must be conducted directly by the authors. AI must not serve as an author or collaborator in the manuscript's content analysis.
|
References
Citing fabricated references generated by AI is prohibited.
Citing a chatbot itself as a primary scientific source is strictly forbidden.
---
**Acceptable Uses of Generative AI**
* **Pre-writing and Preliminary Studies Phase:**
* Initial brainstorming and ideation.
* Summarizing scholarly articles for the author's personal comprehension.
* Utilizing AI as an advanced search engine for literature retrieval.
* **Manuscript Preparation Phase:**
* Proofreading for grammar and spelling errors.
* Enhancing readability, sentence flow, and academic tone.
* Translating text from other languages into English or Persian.
* **Extracting Information from Scientific Texts:**
* Using tools such as ChatGPT, Elicit, SciSpace, or Claude to process PDF files, summarize content, and extract concepts, variables, or outcomes for the purpose of drafting a "Literature Review" is permitted. **This is strictly contingent upon the author's verification of the extracted data and proper citation of the original source.** If the AI retrieves incorrect information (hallucination) that is subsequently included in the manuscript, the author(s) bear full responsibility.
|
