Unauthorized use of AI in studies has increased.
At Lund University, a student was reported to the disciplinary committee after submitting Nazi propaganda during a course on the Holocaust.
“There was a lot of evidence that Chat GPT was used,” says Elisabeth Geevers, director of studies in history.
It was during a course on the history of the Holocaust that a student at Lund University submitted a take-home exam that caused the examiner to react.
Parts of the text resembled Nazi propaganda, and several of the references were fabricated, the examiner believed.
The essay was forwarded to Elisabeth Geevers, director of studies in history, who agreed.
“I spoke to the teacher on the course. He is an expert on Nazi propaganda and immediately realized that there were formulations in the paper that did not follow the scientific debate,” says Elisabeth Geevers.
Anti-Semitic myth in the exam
Among other things, the student reproduced a well-known anti-Semitic conspiracy theory that the Russian Revolution was a Jewish plot – an idea that was later used as an ideological basis in Nazi propaganda.
The teacher suspected that the entire exam was written by AI. Among other things, because the references were not correct and because the essay contained formulations that usually appear on “questionable internet sources,” says Elisabeth Geevers.
– We do not believe that the student wanted to spread this. Instead, we see it as an unfortunate result of using Chat GPT, she says.
The essay was failed and the student was reported to the disciplinary committee.
The student herself denies using AI. She says she was sick for several weeks, missed classes, and submitted a “bad and stressed” answer. However, she admits she made a mistake when she referred to anti-Semitic conspiracy theories.
The honor was dropped.
When the disciplinary committee took up the case, the student had already finished her semester and left Sweden. It is unclear to which country. Since she was no longer a student at that time, the university could not take disciplinary action, and the case was closed.
– I, who made the report, am convinced that there was much evidence that Chat GPT was used.
In 2024, 237 students were suspended or warned for unauthorized use of AI tools at universities and colleges in Sweden. This was an increase from 108 cases in 2023.
At Lund University, the disciplinary committee has handled 67 cases so far in 2025. Of these, 21 cases concerned misleading use of AI. 14 led to suspensions.
