AI systems in teaching

The university, too, is attentively following developments in the field of AI text production. The discussion has intensified after ChatGPT was released, and university members are looking at the opportunities as well as challenges posed by language-based, AI text production systems in the context of teaching, learning and conducting exams. Because the number of such tools and their potential applications as well as the quality of the texts and analyses they produce is only expected to increase, the university is actively looking for meaningful uses of AI in university teaching, but also critically assessing the implications for teaching and exam settings.

Recommendations for the current exam period (spring of 2023)

In response to concerns that cheating attempts in written exams/assignments may increase after ChatGPT has been made freely available and has generated great media attention, the University of Konstanz recommends the following:

• Expect that students have access to AI language models such as ChatGPT or other AI tools, and actively discuss this topic with your students. Raise your students' awareness that many issues involving data protection, copyright law and the reuse of data they input have not yet been resolved.

• Where it is necessary and possible, you can change assignments in written performance assessments, so that AI tools are either less useful or their usage is easier to spot. For further advice, feel free to contact Alexander Klein (University Didactics) for current information about the quality of texts and materials generated by AI.
• As usual, students submit a declaration of independent work assuring that they have not had help from third parties. Especially in the case of work that (also) tests students' ability to formulate thoughts in words, you can explicitly point out to students that, while completing such tasks, the use of AI tools (even for parts or draft versions) is currently considered to be impermissible help from third parties.
• If there are clear grounds for suspecting a student has used an AI language model, the respective Examination Board must summon the student for a hearing on the alleged cheating. If the suspicion is confirmed, the student will be sanctioned with a grade of 5.0 (fail). If this happens more than once, the student can be permanently banned from the study programme.

Declaration of independent work

The goal for the summer semester is to consider the purposes and types of performance assessments for which AI tools can or should be explicitly permitted, and whether or how the use of such tools must be stated, as well as where AI tools should be explicitly prohibited. The working group "AI in teaching" will work on formulating corresponding declarations of independent work for different settings. The working group involves the Division of Legal Affairs as well as the Office for Curricular Affairs and ideally, at least one university teacher from each faculty. You can reach the working group by emailing: KI-Lehre@uni-konstanz.de

Working group AI in teaching

Over the medium to long term, the aim is not only to create suitable declarations of independent work, but also to drive the further didactical development of teaching and exam settings. The working group will address questions such as:

• Where and how do performance assessments need to be adapted (on-campus vs. remote exams, written vs. oral exams, opportunities and limitations of online formats)?
• In the future, how should students be taught and practice (scientific) writing and research?
• How can AI be put to good use in different disciplines?
• Which teaching settings support students in expanding foundational cognitive and (written) language competencies instead of simply relying on the use of AI tools?
• Which interdisciplinary and professional competencies in using AI should students acquire (e.g. in the context of ADILT, transferable skills, teacher education programmes)?

The working group also plans to systematically observe and compile information on developments at German and international universities, to bring together experiences and perspectives from different university members, and to create and share ideas, suggestions and guidelines for teaching. To enable the greatest possible synergies, all university members active in this area are welcome to email KI-Lehre@uni.kn so that the results and experiences can be shared with all departments and teams around the university.

The working group aims to promote networking within the university and the use of materials and results that have been developed elsewhere. Please contact us, too, if you would like to join the working group.