AI for Teaching and Examination
ChatAI – Access
A data protection-compliant access to various LLMs (Large Language Models) is provided via the academic service portal of the State of Lower Saxony, Academic Cloud, here.
The Academic Cloud is provided by the Gesellschaft für wissenschaftliche Datenverarbeitung mbH Göttingen (GWDG). It enables the secure testing and use of AI services, including various open-source and OpenAI models, as well as a RAG system.
Further information about the project can be found here.
Scenarios for Teaching
Generative AI can be used effectively in a variety of ways, including developing digital literacy and data literacy, reflecting on academic practices, simplifying standardized writing tasks, overcoming initial barriers to writing, fostering creativity, and much more.
Its use can be explored in teaching and learning scenarios in line with the objectives and content of the degree programme. Students’ level of competence should be taken into account. It is advisable to communicate transparently the type and extent of AI use (or non-use), reflect on the principles of good academic practice with students, and agree on appropriate forms of documentation. Prof. Dr. Christian Spannagel from Heidelberg University of Education has developed an example of Rules for the Use of ChatGPT and Other Text Generators in Written Assignments: Rules for tools.
The adapted declaration of independent work for the University of Bremen can be found at the Zentrales Prüfungsamt.
On this page, we have compiled various ideas on how AI tools can support lecturers in their professional activities. This collection does not claim to be exhaustive; rather, it is intended to provide inspiration for further exploring the use of AI.
We assume that AI-generated outputs should not be adopted uncritically, but rather treated as initial drafts and carefully reviewed.
Proven Didactic Scenarios
- Didaktische Handreichung zur praktischen Nutzung von KI in der Lehre (2025) v2 by the Working Group on Digital Media and Higher Education Didactics of the German Society for Higher Education Didactics (dghd), in cooperation with the German Informatics Society (GIW) / Society for Media in Science (GMW).
- 101 Creative ideas to use AI in education (2023), coordinated and edited by Nerantzi, C., Abegglen, S., Karatsiori, M. and Martinez-Arboleda, A., who teach and conduct research at the Universities of Calgary, Leeds and Macedonia.
Planning Teaching
- Create semester plans
- Formulate learning objectives
- Develop interactive teaching and learning scenarios
- Create individualized learning materials
- Repurpose existing content in new formats
- Create standardized types of text
Designing Teaching
- Provide individualized support for learning processes
- Support self-directed learning phases
- Provide feedback
- Critically examine AI-generated outputs and compare them with scientifically validated findings
- Support students during group work and break down complex tasks into smaller subtasks
Evaluating Teaching
- Generate exam questions that students can use to prepare for an examination
- Create drafts for course evaluations
- Evaluate teaching concepts
- Check programming code for specific security aspects
Further Information
- KI in Studium und Lehre mit zahlreichen Beispielen auf e-teaching.org
- Gimpel, H., Hall, K. et al.: Whitepaper: „Unlocking the Power of Generative AI Models and Systems like GPT-4 and ChatGPT for Higher Education - A Guide for Students and Lecturers“. University of Hohenheim, March 20, 2023
- Kommentierte Linksammlung des Hochschulforum Digitalisierung:
- https://hochschulforumdigitalisierung.de/de/blog/Hochschullehre-KI-gestuetztes-SchreibenTextgenerierende KI kann sinnvoll eingesetzt werden, unter anderem um Digital Literacy und Data Literacy zu entwickeln, wissenschaftliche Praktiken zu reflektieren, standardisierte Schreibaufgaben zu vereinfachen, Anfangsbarrieren beim Schreiben zu überwinden oder Kreativität zu fördern und vieles mehr.
Forms from the Central Examination Office
The ZPA Forms page provides the following documents, supplemented with information on the use of AI:
- Written assignments – Declaration of Independent Work and Declaration of Consent to Verification Using Plagiarism Detection Software
- Copyright Declaration, Declaration of Consent to the Publication of Bachelor’s/Master’s Theses, and Declaration of Consent to Electronic Plagiarism Checks
- Example Documentation of AI Use in Teaching
Prüfungen KI-sensibel gestalten
To ensure fair assessment in times of generative AI, it is advisable to reflect on the following aspects:
- Adapting assessment criteria
- Adapting assessment tasks
- Strengthening formative learning support
- Strengthening competence-oriented assessment
- Preventive measures against academic misconduct
The following Lernzieltaxonomie (CC BY Hanke 2023) provides recommendations for designing learning and assessment.
Further Suggestions for Designing Assessments
- Formulate topics or questions in ways that encourage critical thinking
- Personal experiences and application to examples from course materials and lectures
- Highly specific and applied topics and questions
- For programming assignments: a combination of code-based and concept-based tasks
- “Authentic assessments” that require students to demonstrate creativity and interdisciplinary skills
- Interviews, discussions, data collection and analysis
- Greater emphasis on the process, rather than solely on the final outcome, in assessment formats such as essays and term papers