AI-generated Policy

AI-generated Policy

The use of generative artificial intelligence (AI) or AI-assisted technologies in the writing process must be approached with caution. AI-generated content must undergo thorough review and careful editing to ensure accuracy, completeness, and objectivity. Any use of such technologies must be explicitly disclosed in the manuscript, and this disclosure will appear in the published version of the work. Full responsibility for the accuracy, integrity, and originality of the manuscript rests entirely with the author(s). (Narongrit Sombatsompop, 2025)

AI Use Policy for Article Writing

With the advancement of Artificial Intelligence (AI) technology and automated content generation tools such as ChatGPT, Claude, Gemini, and other tools, Modern Public Administration Science Journal recognizes the necessity to establish guidelines for using these technologies ethically and responsibly.

This policy aims to provide clear guidance to authors, reviewers, and the editorial board on the appropriate use of AI tools, with consideration for accuracy, credibility, and academic responsibility.

1. General Principles

1.1 Author Responsibility

Authors are primarily responsible for all content

  • Authors must be responsible for the accuracy, completeness, and credibility of all content in the article
  • Use of AI tools does not reduce or transfer this responsibility
  • Authors must verify and confirm the accuracy of all information obtained from AI

1.2 Transparency

Must disclose AI use clearly

  • Authors must disclose which parts of the article used AI tools
  • State the name of AI tools and version (if available)
  • Explain methods and purposes of AI use

1.3 Academic Integrity

Maintain academic standards

  • AI use must not compromise the quality and credibility of academic work
  • Must not use AI to conceal information, distort facts, or create false data
  • Must maintain originality and academic value of the work

2. Acceptable Use of AI

2.1 Technical Assistance

AI is permitted for:

Language and Grammar Improvement

  • Check and correct grammar
  • Improve word choice and sentence structure
  • Check spelling
  • Improve writing clarity

Example:

Acceptable: "Used Grammarly or ChatGPT to check grammar and improve  English writing"

Language Translation

  • Translate articles or abstracts to other languages
  • Must be verified and edited by someone knowledgeable in that language
  • Must state that AI was used for translation

Example:

Acceptable: "English abstract translated using DeepL and verified by an expert"

Formatting and Citations

  • Format references according to APA 7th Edition
  • Format tables and figures
  • Check accuracy of in-text citations

Example:

Acceptable: "Used Zotero or Mendeley to manage references"

2.2 Research Support

AI is permitted for:

Literature Search and Review

  • Search for related research
  • Summarize key points from numerous documents
  • Identify research gaps
  • But must read and verify original documents independently

Example:

Acceptable: "Used AI to summarize main points from 50 research articles, then researcher read and analyzed all original documents"

Data Analysis

  • Help write code for statistical analysis
  • Verify accuracy of calculation formulas
  • Create graphs and charts
  • But must understand and verify entire analysis process

Brainstorming

  • Help think of research topics
  • Suggest preliminary conceptual frameworks
  • Recommend analysis approaches
  • But must develop and refine with own thinking

2.3 Improving Article Quality

AI is permitted for:

Consistency Checking

  • Check content consistency
  • Check completeness of article structure
  • Suggest presentation improvements

Creating Diagrams and Illustrations

  • Use AI to create diagrams, conceptual frameworks, or models
  • Must state created by AI and specify tool
  • Must verify accuracy and appropriateness
  1. Unacceptable Use of AI

3.1 Content Generation

AI is forbidden for:

Writing Entire Articles or Major Sections

  • Forbidden to have AI write introduction, methodology, results, or discussion in their entirety
  • Forbidden to copy text from AI directly without significant modification
  • Forbidden to use AI instead of author's own analysis and synthesis

Reasons:

  • Violates authorship principles
  • AI content may be inaccurate, incomplete, or false
  • Lacks human analytical thinking and expertise

3.2 Fabrication

Strictly forbidden:

Using AI to Create Research Data

  • Forbidden to use AI to create statistical data
  • Forbidden to use AI to create quotes or interview data
  • Forbidden to use AI to create tables, figures, or graphs not derived from actual data

Using AI to Create False References

  • Forbidden to cite research that AI created but does not exist
  • Must verify existence of all reference entries
  • Forbidden to use information from AI that cannot be source-verified

Consequences:

  • Considered serious academic misconduct
  • Article will be immediately rejected or retracted
  • Institution will be notified and permanent submission ban

3.3 Plagiarism

The following actions are forbidden:

Copying from AI Without Citation

  • Copying text from AI is considered plagiarism
  • Even though AI is not human, generated content may come from other sources

Using AI to Evade Plagiarism Detection

  • Forbidden to use AI to modify copied text to evade detection software
  • Considered academic fraud

3.4 AI as Author

Forbidden to list AI as author or co-author

Reasons:

  • AI cannot be responsible for content
  • AI has no understanding or intention in research
  • Does not meet authorship criteria

Unacceptable examples:

Not acceptable: "Somchai Jaidee and ChatGPT (2024)"

Not acceptable: "ChatGPT, Somchai Jaidee (2024)"

Correct method:

Acceptable: State AI use in methodology or acknowledgments

4. Disclosure of AI Use

4.1 Disclosure Requirements

Authors must disclose AI use in the following sections:

In Methodology Section If AI was used in research process, such as:

  • Data analysis
  • Code writing
  • Natural language processing

Example:

"Qualitative data analysis used ChatGPT-4 to assist in preliminary

categorization, then researcher verified and refined categorization

independently"

In Acknowledgments Section If AI assisted in writing or improving the article

Example:

"Authors used Grammarly and ChatGPT-4 to check grammar and improve

English. Authors verified and are responsible for all content"

In Figure Captions If AI was used to create illustrations

Example:

"Figure 1 Research Conceptual Framework (created by DALL-E 3 and modified by researcher)"

4.2 Disclosure Format

Standard Form:

AI Use Statement

Authors used artificial intelligence tools in preparing this article as follows:

  1. Tools used: [specify name and version]
  2. Purpose: [explain what it was used for]
  3. Scope: [specify sections used]
  4. Verification: [explain how verification was done]

Authors verified and are responsible for the accuracy of all content

Example of Good Disclosure:

AI Use Statement

Authors used artificial intelligence tools in preparing this article as follows:

  1. ChatGPT-4 (OpenAI, 2024) was used to check and improve English grammar in the abstract
  1. R Studio with tidyverse packages used ChatGPT-4 to help write and verify R code for statistical data analysis
  1. DeepL Pro was used to translate abstract from Thai to English, with expert verification

Authors verified and are responsible for the accuracy and completeness of all content in the article

5. Guidelines for Reviewers

5.1 Detecting AI Use

Signs of Inappropriate AI Use:

Writing Style

  • Text that is overly formal or perfect
  • Lacks connection to Thai or local context
  • Contains errors characteristic of AI (AI hallucinations)

Content

  • Citations of non-existent documents
  • Inconsistent or contradictory information
  • Lacks depth of analysis

Methodology

  • Methods that are not actually feasible or unreasonable
  • Data that appears fabricated

5.2 Reporting Suspected AI Use

If reviewers suspect inappropriate AI use:

  1. Specify in review report with evidence or reasoning
  2. Notify editorial board for further investigation
  3. Do not immediately reject article just because of AI suspicion

Example report:

"Reviewer notes that article contains multiple reference citations that cannot be verified (e.g., Smith & Johnson, 2023 does not appear in any database), which may indicate AI-generated false references.

Request editorial board verify further"

6. Guidelines for Editors

6.1 Verifying AI Use

Verification Procedures:

  1. Check Disclosure
  • Verify whether authors disclosed AI use
  • Assess appropriateness of disclosed AI use
  1. Content Verification
  • Use AI detection tools (e.g., GPTZero, AI Content Detector)
  • Verify credibility of references
  • Check data consistency
  1. Author Inquiry If doubts arise:
  • Request explanation from authors
  • Request additional evidence (e.g., raw data, code)
  • Give authors opportunity to correct or add disclosure

6.2 Measures for Violations

Violation Levels and Measures:

Minor Level: Failure to Disclose AI Use

  • Measure: Request correction to add disclosure
  • Impact: Does not affect publication

Moderate Level: AI Use Beyond Permitted Scope

  • Measure: Request revision of AI-generated content
  • Impact: Return for revision (Major Revision)

Severe Level: Using AI to Create False Data or False References

  • Measure: Immediate rejection
  • Impact: 2-5 year submission ban and notify institution

7. Case Studies and Examples

7.1 Acceptable AI Use Scenarios

Case 1: Improving English

Situation: Thai researcher writes article in English but unsure of grammar

AI Use:

  • Used Grammarly Premium to check grammar
  • Used ChatGPT-4 to improve sentence structure
  • Author read and revised all suggestions

Disclosure:

"Authors used Grammarly Premium and ChatGPT-4 to check and improve English grammar. Authors verified and are responsible for all content"

Decision: ✅ Acceptable

Case 2: Data Analysis Assistance

Situation: Researcher has large dataset and wants to use Python for analysis

AI Use:

  • Used ChatGPT-4 to help write Python code
  • Tested and verified code independently
  • Attached code as Supplementary Material

Disclosure:

"Researcher used ChatGPT-4 to help write Python code for data analysis (see Supplementary Material). Researcher verified correctness of code and all analysis results"

Decision: ✅ Acceptable

Case 3: Literature Review Assistance

Situation: Hundreds of related research articles, need to summarize trends

AI Use:

  • Used Semantic Scholar AI to find related research
  • Used ChatGPT-4 to summarize main points from 100 abstracts
  • Read 30 important original documents independently
  • Wrote literature review independently by synthesizing readings

Disclosure:

"Researcher used Semantic Scholar and ChatGPT-4 to help summarize preliminary main points from numerous research articles, then read and analyzed all important original documents independently"

Decision: ✅ Acceptable

7.2 Unacceptable Scenarios

Case 4: AI Writing Entire Article

Situation: Author had ChatGPT write entire article from research topic

AI Use:

  • Input "write research article about..."
  • Copied content from AI directly
  • Made only minor edits

Decision: ❌ Not acceptable - violates authorship principles

Case 5: AI-Generated References

Situation: Author asked ChatGPT "what research exists about..." and cited as AI suggested

AI Use:

  • AI recommended 10 research articles
  • Author cited without verifying they exist
  • 5 articles do not actually exist (AI hallucination)

Decision: ❌ Not acceptable - considered serious data fabrication

Case 6: Using AI to Create Research Data

Situation: Researcher lacked time to collect data, had AI create numerical data

AI Use:

  • Asked AI to create plausible statistical data
  • Used data for analysis and report writing

Decision: ❌ Not acceptable - considered data falsification, permanent ban

8. Tools and Resources

8.1 AI Detection Tools

For Editorial Board:

Note: These tools are not perfect; must be used with human verification

8.2 Recommended AI Tools

Language Improvement:

  • Grammarly
  • QuillBot
  • LanguageTool

Translation:

  • DeepL
  • Google Translate (for preliminary translation)

Reference Management:

  • Zotero
  • Mendeley
  • EndNote

Data Analysis:

  • ChatGPT/Claude (for code writing assistance)
  • GitHub Copilot

9. Frequently Asked Questions (FAQs)

Q1: Can I use ChatGPT to help write the introduction?

A: Not recommended. The introduction should be the author's own thinking and analysis. You may use AI for brainstorming or language improvement, but must write main content yourself.

Q2: If I use AI to help write data analysis code, must I disclose it?

A: Yes. Should disclose in methodology section and attach code as Supplementary Material.

Q3: I used Grammarly to correct grammar, must I disclose it?

A: Recommended to disclose in acknowledgments, even if basic use.

Q4: AI created references that don't exist, what should I do?

A: Absolutely forbidden. Must verify existence of all reference entries. Citing false references is serious misconduct.

Q5: I translated abstract with DeepL, must I disclose it?

A: Yes. Should state DeepL was used for translation with expert verification.

Q6: What happens if reviewers suspect AI use?

A: Editorial board will investigate and request explanation. You must be ready to show evidence and explain work process.

Q7: Can I use AI to create illustrations?

A: Yes, but must state in figure caption that it was created with AI, which tool, and verify appropriateness of image.

Q8: Will the journal change this policy in the future?

A: This policy will be regularly reviewed according to AI technology advancement.

10. Enforcement and Review

10.1 Enforcement

This policy applies to:

  • All articles submitted from January 1, 2026
  • All authors, reviewers, and editorial board members
  • All article types (research articles, academic articles)

10.2 Review and Updates

  • Policy will be reviewed every 6 months
  • Updated according to AI technology advancement
  • Academic community feedback welcomed
  • Updates published on journal website

11. Contact and Consultation

For questions about AI use policy, contact:

Email: jppissues@gmail.com or dpi.fms@snru.ac.th
Subject: AI Policy Inquiry
Phone: 081 596 5492

Editorial board welcomes consultation before article submission

Conclusion

Modern Public Administration Science Journal recognizes the potential of AI technology in assisting research and article writing. However, we believe that:

Human thinking and analysis are the heart of academic work
AI is an assistance tool, not the main content creator
Transparency and responsibility are essential
Authors must be responsible for all content

Ethical AI use will help increase work efficiency while maintaining academic standards and value.

Effective Date: January 1, 2026
Last Review: January 1, 2026
Next Review: January 1, 2027