34th Open Science Meet-up: Challenges of AI for (Open) Science and Education

Monday 22 Sept 2025, 14:00 → 16:00 Europe/Berlin

https://uni-goettingen.zoom-x.de/j/61443464097?pwd=WPGDxIRchwU6490rP2DNvkbf5eauHe.1 (virtuell)

Astrid Orth (SUB Göttingen), Birgit Schmidt (SUB Göttingen), Daniel Beucke (SUB Göttingen), Julika Mimkes (SUB Göttingen)

Benefits but also several challenges can play out when applying AI in higher education and research.

Unintended consequences can arise from the use of (generative or other) AI-assisted tools and practices on teaching and research, such as sloppy writing and (undetectable) plagiarism. Concerns have grown regarding effects on students’ learning if substantial parts of intellectual work are outsourced to large language models (LLMs).

Similarly, one can ask what constitutes responsible use of AI in publishing and reviewing, and several publishers have created guidance for authors (including approaches for AI acknowledgement statements).

Adverse effects on information and data infrastructures have also emerged, in particular, as the creation of AI models heavily depends on access to large datasets as training data. Openly licensed collections and corpora are frequently used for this purpose. In particular, Open Access repositories, journals and platforms are targeted and experience a increasingly challenging load of requests by AI crawlers.

In this session, we would like to share and discuss observations and experiences, possible challenges and adverse effects and how these might be managed or mitigated.

Pad: https://pad.gwdg.de/B__qfflFSrSqhv_Sxw3wuA

14:00 → 14:15 Welcome & Round of introductions

14:15 → 15:15 Short presentations

• Open access repositories as a target: Results from COAR’s AI bot survey - Daniel
• Effects on teaching and learning - Angelika Thielsch
• Use of AI in academic writing - Ella Grieshammer
• Challenges and pitfalls in the development of AI-powered services and infrastructures - Dominik Meier

Questions to consider: what kind of AI methodologies or tools are used, what are intended purposes, what are general or possible shortcomings or challenges, what guidance is available (or missing), etc.

15:15 → 15:45 Discussion

15:45 → 16:00 Summary, updates & what’s next