Recently, a researcher at Lund University got reprimanded after their doctoral thesis was flagged for AI use. The researcher, who claimed to not have used any AI tools, is allowed to keep the doctoral title despite the conviction for research misconduct.
Doctoral research projects are known to be subjected to rigorous academic standards as they are meant to prepare doctoral researchers for a future in academia. The use of AI in research processes is a contested topic, raising questions regarding academic rigor, integrity, and scientific ethics.
This February, Universitetsläraren reported that the National Board for Assessment of Research Misconduct (NPOF) had flagged a Lund University doctoral thesis for AI use, stating the project contained 14 fabricated references. The project’s reference list contained several non-existent sources, miscited titles, and incorrectly referenced DOI’s. As the thesis had already been defended, and approved by the supervisor as well as the examining committee, the researcher is allowed to retain the doctoral title.
A closer look into the NPOF-assessment shows that the fabricated references only appeared in the reference list, and all the in-text references were correct. According to the report, the doctoral researcher posited that because all reference mistakes merely occurred in this reference list, there was no forgery nor plagiarism.
The incorrect references in the list were the result of stress during the doctoral research process, and the use of reference tools for the compilation of the bibliography. Furthermore, the researcher stated that a correct list has been submitted to the committee, and that the ultimate aim had been to cite female authors in a male-dominated field.
While the use of AI by itself is not conclusive to determine research misconduct, the case highlights the increasing role AI-tools play in the research process, even within doctoral spheres. As of today, there exists little research that assesses the role of AI in doctoral research processes.
The University of Linköping is currently conducting an interview study on the views of doctoral students regarding the role of AI in thesis work, but findings are yet to be published. Meanwhile, students continue to be reprimanded for the use of AI in assignments, although reports of AI cheating at Lund University are relatively low in comparison to other Swedish universities.
The use of AI in university work remains a subject of academic debate. In the meantime, Lund University has published its own guidelines for the use of artificial intelligence in the research process. In one of the university’s own news publications, pro-vice chancellors Jimmie Kristensson and Per Mickwitz establish that: ”The starting point when using AI in research is to follow the basic principles of research ethics and let these guide you regarding how you are to act in your capacity as a researcher.”
The university furthermore refers to research guidelines published by the Swedish Research Council, encouraging researchers to be guided by the principles of reliability, honesty, respect and responsibility when using AI tools.
