Research Training Committee policy statement on ChatGPT
At its ordinary meeting on 21 March 2023, the Research Training Committee at Uppsala University’s Disciplinary Domain of Medicine and Pharmacy (Swedish: KUF) discussed the role that generative AI – such as the popular service ChatGPT – may play in the higher education sector in the future.
The Committee emphasises that doctoral students must continue to be trained in the craft of scholarly publishing, methods, logic, ethics and principles in addition to the actual writing process – but at the same time wishes to shift the focus from the actual generation of a text to responsibility for its content, context and meaning. The responsibility for this training lies with the student, their supervisor and research team.
Consequently, the Committee believes that generative AI should be seen as a natural step in the development of various computer-assisted tools for text and image generation. This trajectory began with spellcheckers and electronic thesauruses, and in recent decades, tools for syntax checking and grammar support have become natural complements to text generation from pre-school onwards.
It is therefore not forbidden, or even discouraged, to use generative AI for work – instead, researchers and graduate students are encouraged to take advantage of the opportunities provided and train themselves to use the tools in the best possible way. However, they must take responsibility for the content of the text, as this cannot be blamed on AI.
For example, the use of generative AI could be formulated in the following way in the introductory chapter of compilation theses:
This thesis was partially created with the assistance of GPT-4, Open AI’s large-scale language generation model. Upon generating draft language, I reviewed, edited, and refined the text to my own liking and retain full responsibility for the content presented in this publication.
Peter Stålberg, Professor, Vice-Dean, and Chair of the Research Training Committee