AI at universities: revolution or risk for exams?

AI at universities: revolution or risk for exams?

Nowadays, when artificial intelligence (AI) is finding its way into many areas of life, universities are intensively discussing how this technology affects teaching and examinations. Henrik te Heesen, professor at Trier University in the Department of Environmental Planning and Technology, sees AI not only as challenges but also as enormous opportunities for students and teachers. In an interview with the SWR He emphasizes that traditional exams are no longer appropriate in the era of AI and warns that an “AI two-class society” could emerge among students.

Te Heesen highlights that AI can act as a learning support by providing students with individual explanations and appropriate practice tasks. This support has the potential to revolutionize the learning process. In addition, model exams can be generated with AI, which was previously done manually by student councils. This not only means time savings, but also greater efficiency. He estimates that AI can do about 80% of the work, with the rest requiring manual post-processing.

Opportunities and challenges of AI in teaching

But alongside the positives, there are also challenges, particularly around the cost of powerful AI programs, which may pose a problem for some students. The Rhineland-Palatinate Virtual Campus offers a helpful solution here by providing free AI applications for students.

In the discussion about AI as a touchstone for university teaching, a deeper understanding of educational and socialization processes is also required. The contribution by Prof. Dr. Boris Zizek and Bettina Gautel University Forum for Digitalization addresses the need to sensitize students to reflective actions and democratic values. Principles for using AI tools in teaching should be formulated and criteria for good scientific practice should be clearly communicated.

The challenges surrounding AI in exams, especially cheating, should also not be underestimated. Term papers and exams could be under pressure because AI-generated texts cannot be easily detected using conventional plagiarism software. Therefore, universities have put the development of new examination formats, including oral examinations, on the agenda.

The path to an AI-supported future

Te Heesen sees training students in using AI tools as essential. In order to use the technologies sensibly, an understanding of their risks and opportunities is necessary. Teachers should also make it transparent which AI tools are permitted in exams. The aim is to use the AI ​​tools as a “sparring partner” for critical thinking and to promote reflection processes in exam design.

In the constantly evolving educational landscape, it is crucial to adapt to new circumstances. This is the only way to exploit the full potential of artificial intelligence, which will not only sustainably improve teaching, but also the entire examination and educational culture.