Werkzeugkasten Digitale Lehre
Reiter
AI Chat
What is the AI Chat ?
In the “AI Chat” section, you will find training materials and direct access to various generative AI models and LLMs. The University of Freiburg wants to optimally prepare students and employees for the use of AI tools and promote digital competence, transferability as well as critical thinking and judgment when using AI. This introduction and supplementary learning modules are intended to support the integration of AI as a valuable tool in everyday working, teaching and studying life. With the help of the AI Chats, you currently have the opportunity to try out various Large Language Models (LLMs) in ILIAS as part of a pilot project. In addition to ChatGPT (via an API access of the university), we provide you with various models that are operated here in the data center on university hardware (llama3.1, gemma2, mistral-nemotron etc.).
Participation in the pilot is voluntary. However, participation in the mandatory initial survey is a prerequisite for access to the AI chats.
If you have any questions, please contact us by e-mail at ki@rz.uni-freiburg.de
What can the AI Chat be used for?
For many people, AI tools are already an integral part of their daily working environment. The piloting of the AI chat in ILIAS offers various opportunities to try out AI models in compliance with data protection regulations. Among other things, teachers will be able to easily and securely integrate an AI-supported chatbot into their teaching environment in ILIAS in order to efficiently create teaching materials, personalize learning processes and promote digital skills among learners.
In order to optimally prepare you for the use of AI tools, we offer two short online training courses that you can complete independently. Part 1 offers a brief introduction to the topic of AI, provides information on data protection and the responsible use of generative AI in accordance with the EU-AI Act. The second part of the learning module explains the opportunities and challenges arising from the use of generative AI in higher education.
Example
AI-supported chat assistance: Targeted system prompts are used to prepare the AI assistant to support learners with various tasks, such as helping them to understand certain content or to reflect in greater depth.
The AI Study Buddy, for example, is designed to support your students as an individual learning companion. The aim is to help students structure their learning process and offer personal support. It should help with the creation of study plans and promote a deeper understanding through targeted questions. Instead of giving direct answers, the AI Study Buddy is designed to encourage students to look for solutions themselves and help identify potential gaps in their knowledge. In this way, generative AI is intended to enrich the learning experience and strengthen skills.
The chatbot is preconfigured with the following system prompt and responds to each of your prompts according to this specification to help students optimize the prompts for different use cases:
"You are a PhD student helping others in lower semesters learn for their exams. You will be provided a topic or course title with an optional description of a specific topic and you will help and guide the person asking to a deeper understanding. You can assume a technical background and a higher-level education. Try to design a study plan and provide actionable task, if asked. Compile a list of questions that might come up in an exam and ask follow-up questions if the answers are not precise or correct enough. Do not give the correct answers to these questions right-away but wait for the student to aks for the answers explicitely. Try to anticipate the level of understanding required by the outlined course, if unsure, ask at what level the course is. Be polite and most of all, helpful. Always answer in the language that the user has used."
Important note/Please note:
When using AI in administration, research, teaching and studies, the protection of personal data is crucial in order to protect the privacy of the individual and the privacy of those affected. Only AI systems approved by the University of Freiburg may be used in an official context. The input of personal data or other sensitive information into generative AI systems provided by the university is not permitted.
Help and Support
Do you need further information, help or support?
If you have any questions about ILIAS, please contact us ...
- via our forum. There you will find many questions about ILIAS that have already been answered, and you have the opportunity to ask questions yourself.
- via the e-mail address ki@rz.uni-freiburg.de.
In all other cases, please contact us by e-mail at elearning@rz.uni-freiburg.de.

Further use as OER explicitly permitted: This work and its contents are - unless otherwise stated - licensed under CC BY-SA 4.0. Please attribute according to TASLL rule as follows: "Digital Teaching Toolbox" from E-Learning Department, University of Freiburg, Licence: CC BY-SA 4.0
Zuletzt geändert: 4. Juni 2025, 11:33, [lw1125@uni-freiburg.de]