Generative AI tools such as ChatGPT became widely available in autumn 2022. In February the following year, a student at Uppsala University admitted using ChatGPT to help with written assignments. It was the first Swedish case where a student received a warning for cheating with the help of AI, something which Universitetsläraren wrote about at the time.

Since then, a survey by Universitetsläraren of 35 universities and colleges shows, 770 cases of AI cheating have resulted in suspension. So far this year, students have been suspended for cheating related to some form of unauthorised use of AI tools in 467 cases, up from 207 last year. In addition, 75 warnings have been issued since the beginning of 2023.

At Stockholm University, students have been suspended in 77 cases, the highest number of any higher education institution. Additionally, the university has issued 24 warnings in the past three years.
“We are Sweden’s largest higher education institution for first- and second-cycle courses and programmes, so from that perspective we would be expected to be at the top of the list,” says Clas Hättestrand, chair of the disciplinary committee at Stockholm University.
“We have also been quite active in raising awareness among our teachers about the risk of cheating through the use of AI. It is possible that this has also had an effect.”

He describes how staff are actively informed about AI use. Guides, courses and workshops have also been developed.
“I have noticed that there is a great deal of interest in these issues and a willingness to discuss them. This applies to both the positive aspects of AI, which, for example, our higher education pedagogy unit works with, but also, of course, how we can work with examination methods,” he says.

Linköping University has had 63 cases of students being suspended for unauthorised use of AI tools in the past three years, with 47 of those occurring in 2025. In addition, 12 warnings have been issued.

The increase in suspensions over time is partly due to more and more students using AI in unauthorised ways, according to Karin Axelsson, pro-vice-chancellor at the university.
“I also think that teachers have become better at explaining what is and is not allowed,” she adds.

The introduction of generative AI in higher education, she believes, has also made university teachers more aware of cheating, and perhaps more likely to report it.
“As a teacher, you read and think about submitted texts in a different way than you did before.”

She is not surprised that the curve for suspended students is trending upwards.
“We are doing quite a lot right now at the university that should mean that it, (the curve, editor’s note), will eventually not only level out but also turn downwards. The current level should be a hump. The vice-chancellor has just announced an action plan to AI-proof all our education.”

Under this action plan, examination methods in all courses and programmes are to be reviewed. SEK 4 million per year has also been earmarked for ‘AI ambassadors’ at each department over the next two years.

Axelsson emphasises that this is not just an action plan to curb cheating.
“We are going to assess the examination methods and procedures for all our courses so that we do not go down the same old paths, for example, with a home exam that has looked the same for years. But we are looking just as much, or perhaps even more, at how we may need to change course content so that we prepare students for a future working life where they can expect generative AI to be used routinely,” she says.