genAI guidelines
Preamble
Using OpenAi ChatGPT 4.0, I created a set of guidelines of what I believe students are allowed to do with genAI and what is prohibited. This is obviously work in progress and changes might be necessary as time goes by.
The guiding principles behind these guidelines are intellectual ownership, accountability, and responsibility.
As a general guideline, Generative Artificial Intelligence (genAI) or Large Language Models (LLM) may only serve as a language or thinking aid, never as a surrogate for scholarly authorship or reasoning.
Guidelines
1. Literature Review and Theoretical Framework
Allowed:
- Discovering relevant keywords.
- Exploring possible search strategies.
- Brainstorming relevant topics, theories, schools of thought, or authors.
- Explaining complex concepts.
Prohibited:
- Writing the literature review or background section.
- Synthesizing and comparing sources without direct engagement.
2. Study Design and Methodology
Allowed:
- Exploring different research designs or data collection strategies.
- Asking for pros and cons of methods or instruments.
- Refining research questions or hypotheses.
Prohibited:
- Relying on genAI to select methodology without student understanding.
- Submitting AI-generated mock designs without adaptation by the student.
3. Data Analysis
Allowed (Quantitative Research)
- Exploring data analytic strategies.
- Generating example code.
- Data annotation (e.g., Part of Speech annotation)
- Creating mock datasets for practice and understanding.
- Assistance with the interpretation of code output.
- Explaining statistical tests/models
- Debugging code
Prohibited (Quantitative Research)
- Submitting analyses performed by genAI without verification.
- Using genAI to interpret findings or draw conclusions without understanding.
- Presenting statistical output without understanding its origin.
Allowed (Qualitative Research)1
- Organizing field notes
- Suggesting preliminary codes or patterns
- Exploring thematic clusters
- Drafting, editing, and tailoring questions or statements (survey/interview/focus groups)
- Transcribing raw data (e.g., with Whisper)
Prohibited (Qualitative Research)
- Creating fake data (interview transcripts, fictional quotes, or made-up ethnographic observations)
- Letting genAI interpret participant narratives without human oversight
- Drawing conclusions or coding without researcher validation
- Uploading sensitive transcripts to AI platforms without proper anonymization or consent
- Use genAI when working with personal or protected data.
- Presenting genAI summaries or visuals as original researcher insights
4. GenAI Use in Academic Writing
The guiding principle here is that students may only use genAI to correct or edit their own writings.
Allowed
- Editing and proofreading: grammar, spelling, punctuation, and word choice.
- Stylistic refinement: improving clarity, tone, and register
- Reordering content for readability or flow.
Prohibited
- Automated drafting and ghostwriting of full paragraphs or chapters.
Checklist for students
- I used genAI to explore keywords or source suggestions.
- I read and evaluated all sources myself.
- I wrote the literature review in my own words.
- I used genAI only to clarify difficult terms or theories.
- I wrote the theoretical background and conceptual explanations myself.
- I understand and can explain the theories or models I used.
- I used genAI only to brainstorm or compare possible research designs.
- I can explain why I chose a particular method or design.
- I used genAI to understand or debug data analysis code or concepts.
- I did not rely on genAI to perform the full analysis without review.
- I can justify every step of the analysis and explain the results.
- Any genAI-generated code was tested, modified, and understood by me.
- I wrote the results and discussion section myself.
- genAI only helped me edit or improve spelling, grammar, or clarity.
- I revised all genAI suggestions to ensure they reflect my thinking.
- I disclosed my use of genAI and retained records of genAI prompts and outputs.
- I understand everything in my thesis and can defend it independently.
Footnotes
Qualitative data that involves personal data requires special care. The use of personal data should comply with GDPR.↩︎
Citation
@online{de_cuypere2026,
author = {De Cuypere, Ludovic},
title = {genAI Guidelines},
date = {2026-04-30},
url = {https://ludovicdecuypere.github.io/posts/2026-04-30-genAI_guidelines/},
langid = {en}
}