Support for interpretation regarding the use of generative AI in doctoral education
Uppsala University's Guidelines for the use of generative AI in teaching and examination (approved by the Vice-Chancellor on November 26, 2024, UFV 2023/2129) encourage teachers and students to explore together the opportunities and challenges that the technology brings in a positive spirit. The guidelines stipulate that if the use of generative AI is restricted in teaching and/or examination, this must be clearly stated in written information. The restriction must be justified on the basis of learning objectives, the nature of the task, or pedagogical considerations. In teaching and examination, it must also be clearly stated in written information whether the use of generative AI is to be reported.
The Faculty of Science and Technology's regulations for the use of generative AI are included in the faculty-wide sections of the general study syllabus to which each doctoral student is admitted. Within the Faculty of Science and Technology, the use of generative AI in doctoral education is generally encouraged, and generative AI tools should be available for use by doctoral students. The purpose of the Faculty of Science and Technology's regulations is to ensure that the degree learning outcomes are met and to provide confidence for doctoral students in their use of generative AI. The regulations apply to all material which is regarded as the doctoral student's own texts and which are therefore examined as texts written by the student themselves. The content of publications, which might have several authors, is naturally also examined at a public defence, but with regard to scientific content rather than how the text is generated. According to the regulations, a statement on the use of generative AI must be submitted.
An example of how to formulate a statement in an examined text is:
- This text/thesis/summary was originally written by me as doctoral student. Generative AI has been used for text translation/linguistic changes/... using the xx tool (version yy). This usage follows the guidlelines for use of generative AI in the faculty's general study syllabus for doctoral education.
Contents of the regulations
The Faculty's regulations emphasize the doctoral student's responsibility for their material and that all material must be produced in accordance with the principles of academic integrity (see references below). In order to achieve the degree learning outcomes of being able to present and discuss in writing, demonstrate intellectual independence, and critically identify and formulate questions (objectives A2, B1, B2, B3, B4, C1, Högskoleförordning (1993:100) | Sveriges riksdag), doctoral students must themselves formulate and develop a first version of a coherent section of text without generative AI. This could, for example, be a chapter in the summary section of the thesis (the cover).
Doctoral students must report their use of generative AI and the tools used, as well as the version, if necessary, to understand what the tool can do. Since it is difficult to control the use of generative AI, a more continuous and discursive examination of the doctoral student's work may be required to monitor the progression of their analytical skills. The doctoral student needs to use generative AI wisely so that a final product is generated that the doctoral student understands and can defend. Generative AI can never be blamed or listed as a co-author.
The regulations can be found in the general study plan
Faculty-wide parts of the general study plan with regulations for the use of generative AI will apply from March 1, 2026.
Faculty-wide parts also apply to doctoral students who began their doctoral education before March 1, 2026, unless they would result in poorer conditions for doctoral students. In addition, there may be subject-specific restrictions on the use of generative AI, as decided by the Doctoral Education Board. There should be no additional regulations outside the general study plan, as this would make it unclear to doctoral students what applies. Similarly, a member of the grading committee should only need to read the general study plan to understand which regulations apply. Any subject-specific restrictions on generative AI must be clearly justified on the basis of pedagogical considerations, degree objectives, and subject-specific objectives within the doctoral education subject.
When deciding on subject-specific parts of the general study plan, the Doctoral Education Board will take into account that large differences between the restrictions risk hindering collaboration across subject boundaries.
Discuss together
The doctoral student, supervisor, and FUAP need to discuss together in which situations it may be appropriate or more problematic to use generative AI, as well as what material may be examined, both in the near future and later on at a public defence. It is advisable to include an interpretation of the regulations (as specified in the general study syllabus) on how generative AI can be used for each project in which the doctoral student participates in the individual study plan. In order to support this discussion, questions for doctoral students, supervisors, and FUAP to consider are listed below:
- In which parts of the doctoral student's project is there scope for the use of generative AI (formulation of research questions, choice of methods, literature searches, data analysis, writing)?
- How can academic integrity and a critical approach be ensured, in the same way as for all material produced during doctoral education? See literature below.
- When is a first version of a coherent section of text which is to be examined with regard to written presentation ready for possible processing using generative AI?
- When is it reasonable, for example, to use generative AI to summarize articles (by other research groups) from a literature search?
- When and how is it appropriate to use generative AI for figures and illustrations? The above regulations mean that it is permissible to use generative AI for the visualization of material to be included in the thesis, for example for illustrations of overall processes.
- When and how is it appropriate to use generative AI for code generation? The above regulations mean that code may be produced using generative AI.
- Do the regulations for the use of generative AI at the faculty differ from those at other collaborating universities, or from the guidelines of relevant journals, in terms of use and/or reporting?
- How should co-authored publications be approached and used in the writing of texts which are to be examined, and how should this be reported?
- What ethical challenges might there be in using generative AI, such as the risk of unintentional plagiarism? See reference Bjelobaba et al 2025.
Literature and other resources
Support and Professional Development – Teaching – Staff – Uppsala University
Tools for good research practice – Research – Staff – Uppsala University
European Code of Conduct for Research Integrity – Revised Edition 2023