Why is it urgent to talk about ethics and AI in academic theses?
Tools like ChatGPT, Copilot, and Gemini are already transforming how students write, research, and produce knowledge. However, the use of these technologies in undergraduate theses raises an essential question: to what extent is it ethical to let a machine participate in the authorship of an academic work?In this post, we will discuss the main ethical dilemmas, best practices, and show which Brazilian institutions are already regulating the use of AI in academic activities.
Main Ethical Dilemmas in Using AI for Academic Theses
- Authorship and responsibility
If part of the text is written by AI, who is the true author? Can the student take full responsibility for something they did not entirely write? - Transparency
The use of AI must be disclosed. Hiding prompts or generated excerpts can be interpreted as a lack of academic honesty. - Accuracy and reliability
AI models may invent data, distort concepts, or create fake references. The student must critically review everything that is generated. - Inequality of access
Not everyone has the same tools, which can create unfair advantages among students. - Learning and autonomy
If AI does the hardest part, is the student truly learning? The line between assistance and dependence is thin. - Plagiarism and originality
Even without copying from another author, using AI-generated text without attribution can still be considered plagiarism.
Best Practices for Ethical AI Use
- Communicate with your advisor: before using AI, agree on whether and how it can be applied.
- Keep records: document the prompts used and describe the tools in your methodology section.
- Review everything: ensure the final content reflects your style, ideas, and reasoning.
- Be transparent: mention in the paper that you used generative AI as support.
- Value human learning: use AI as a support tool, not as a substitute for research.
Universities and Institutions with AI Policies
Several Brazilian institutions are leading the way in creating ethical guidelines for the use of generative AI in academic work. Here are some examples:
| Institution | Document / Policy | Key Points |
|---|---|---|
| SENAI CIMATEC (BA) | Guidelines on Ethical Use of Generative AI | Transparency, privacy, human-centered focus, and requirement to disclose prompts used. |
| Unicentro (PR) | Guidelines for AI Use in Research (2025) | Encourages responsible, ethical, and transparent use; requires AI usage records. |
| Unicamp / UFPR / UFPE / UFG | Guide “Ethical and Responsible Use of Generative AI” | Recommends disclosure of tools used and prohibits AI as co-author. |
| USP – Comunicação & Educação Journal | AI Use Policy in Publications | AI cannot be listed as an author; its use must be described in the manuscript. |
| Saúde e Pesquisa Journal (Unicesumar) | Policy on Generative AI Use | Allows AI for language review only; prohibits generating scientific content. |
| Faculdade BP | Guidelines for Responsible AI Use (2025) | Addresses AI in higher education focusing on ethics, transparency, and accountability. |
| STHEM Brasil | Guide “Ethics and Responsibility in AI Use in Higher Education” | Focuses on institutional ethics, faculty training, and pedagogical best practices. |
How Institutions Can Prepare
Even where there are no specific rules yet, colleges and universities can adopt institutional policies based on four key pillars:
- Transparency: require AI use to be described in the paper.
- Clear boundaries: allow AI only for support tasks (review, formatting, verification).
- Faculty supervision: guide and monitor AI use.
- Critical education: train students and professors to understand the potential and risks of AI.
The International Perspective
Organizations like UNESCO and the OECD have also been developing guidelines for the ethical use of AI in education, advocating for principles of transparency, fairness, privacy, and accountability. These global references can inspire more consistent national policies.
Conclusion
Artificial intelligence is not an enemy of scientific production — it is a powerful tool that must be used with awareness and ethics. An undergraduate thesis should remain an expression of the student’s analytical and creative abilities, with AI acting only as an assistant, never as an author.
Reflecting before using is the first step to using well.