AI Policy

As a group that is researching AI, we are excited by the opportunities of this new technology. However, our experience has shown that relying too much on generative AI in conducting research projects is counterproductive to the outcome. The following rules are binding for all students of the lab:

• We strongly encourage the use of AI to check both writing and code for errors. In practice, this means that AFTER you have written text or code, ask the AI to check for errors, logic and scientific rigour, as well as the appropriateness of citations.
• We weakly encourage the use of AI for finding relevant literature, but note that this should only complement your traditional literature research via search engines such as Google Scholar or just by browsing the reference lists of papers that you read! Note that the sole use of AI for literature research often produces inferior results to traditional research, as AI tools are still not able to reliably understand the importance of the papers.
• We forbid the use of AI to produce text and code from scratch - this has two reasons: 1) the whole point of going to university is that you learn to code and write - there is a good amount of research showing that offloading this task to an AI reduces learning 2) AI written code and text is still often bulky, full of redundancies and without human guidance larger projects deviate from what was originally intended. We warn that the use of AI for these purposes is obvious and may lead to a poor mark for your projects / thesis.
• The use of AI tools (which tool and how it was used) should be disclosed at the end of all theses or projects.

On top of that, we ask students to note the AI policy of our faculty (externer Link, öffnet neues Fenster).