For two years, the GIUB has been addressing the question of what advances in artificial intelligence mean for research and teaching in geography. The ‘AI Working Group’ established for this purpose, consisting of students, research assistants and professors, first drafted a guide for dealing with AI in teaching. After this guide was approved by the board, the working group is now working to provide information on current developments in the field of AI and to stimulate discussion on this topic.
For this reason, AI was the topic of two lecture events on 26 November 2025: the use of AI to predict extreme events and the influence of AI tools on research and teaching in human geography.
In the Lunch Lecture, Prof. Jürgen Gall from the Institute of Computer Science at the University of Bonn reported on the use of AI to predict extreme events. In his presentation, he introduced three application examples from his own work. In the first example, he showed how AI is used to optimise the prediction of forest fires. This involves short-term predictions of spatial events based on dynamic variables. Among other things, the model presented is used to position rescue workers ideally in the area. An AI-driven model developed by his research group that focuses on long-term forecasts deals with the prediction of droughts. This models the probability of droughts occurring in the next two years. A third application example showed how AI-supported models are used to predict flooding events on a global scale. A key conclusion of the presentation was that AI-supported spatial models provide very good forecasts, especially when long-term training data is available.
During the evening event, three panellists discussed how AI tools influence research and teaching in human geography. Dr Johannes Dittmann (Bonn) moderated the exchange between the two external guests, Prof. Dr Marc Boeckler (Frankfurt) and Prof. Dr Veronika Cummings (Mainz), and Dr Jonas Birke (Bonn). In his contribution, Marc Boeckler emphasised that AI is fundamentally changing the production of scientific texts in particular – authors are now becoming more like designers of texts. At the same time, AI remains incapable of collecting data using ethnographic methods. In her contributions, Veronika Cummings addressed primarily research ethics issues relating to the use of AI. A key question is that of responsibility and accountability when AI is used in research processes to accelerate and rationalise processes. Jonas Birke pointed out that AI also raises new questions for human geography as a field of research, such as the change in social interactions or the change in spatial practices through AI.
All three speakers agreed that AI poses a significant challenge for teaching. The main task here is to guide students in their use of AI. While coursework and examination requirements need to be adapted, it is also important to place the necessary trust in students that they will not take easy and reckless ‘shortcuts’. In addition to enabling informed use, it was also emphasised that fair access to premium versions of AI tools should be created.
Following the evening event, the evening came to a close with further discussions in the Red Hall. Here, the student council had organised drinks, snacks and music. The AI Day at GIUB did not provide any definitive answers, but it did provide numerous ideas for the ongoing discussion of what AI means for geography.
