AICSE2026
Workshop: Artificial Intelligence in Computer Science Education (DELFI2026)
Advances in Artificial Intelligence (AI) and Large Language Models (LLMs) are transforming teaching and learning across diverse educational contexts, creating new possibilities for classroom applications, personalized learning systems, and learning analytics that enhance traditional learning methods. In particular, a large impact is notable in computer science (CS) education due to the capabilities of LLMs to generate code. These changes demand a reevaluation of the competencies that need to be fostered in CS. At the same time, AI systems offer many opportunities to expand already existing teaching strategies and approaches. However, there are still many concerns that need to be addressed (e.g., security, quality, reliability). This workshop aims to facilitate exchange among researchers investigating CS education from diverse perspectives to discuss future directions for CS education.
Date: 15.09.2026, 14:00-17:30.
Place: Potsdam, Germany, as part of the DELFI2026.
Contact: aicse@lists.techfak.uni-bielefeld.de.
Main Infos
Tentative Schedule
Here you can find the tentative schedule of the workshop on 15.09.2026.
- 14:00-14:10 Introduction
- 14:10-14:40 Keynote. Speaker: Niels Pinkwart. Topic: Feedback in Programming Education - the past, present and future.
- 14:40-15:30 Interactive Part.
- 15:30-16:00 Coffee Break.
- 16:00-17:20 Spotlight Talks.
- 17:20-17:30 Workshop conclusion.
Spotlight Talks (16:00-17:20)
16:00-16:20 Using Generative AI in Database Education: An Experience Report
Presenter: Andreas Thor
Abstract: This experience report describes three scenarios for using generative AI in database education. We use AI for self-study, for classroom interaction with students, and for creating learning materials. For each scenario, we present two concrete examples and report on experiences from our database courses in recent semesters.
16:20-16:40 Generate First, Verify Later? Students' Use and Validation of GenAI in Software Projects
Presenter: Jan Haas
Abstract: Generative AI (GenAI) tools have become an integral part of students’ programming workflows. However, how students verify AI-generated output in software projects remains largely unexplored. To address this gap, we surveyed 35 students in a 15-week capstone programming practicum at a German university and qualitatively analyzed their open-ended responses. Results show that students used GenAI primarily for code generation, debugging, and concept explanation, and validated outputs mainly through prior review, practical testing, or GenAI itself. Students also expressed a desire to improve their prompting strategies, to use AI more reflectively, and to become better at programming.
16:40-17:00 Herausforderungen bei der LLM-basierten Lückentexterstellung in der Programmierausbildung
Presenter: Patrick Weber
Abstract: In dieser Arbeit wird untersucht, ob mithilfe von Large Language Models Dropdown- Lückentexte generiert werden können, die Studierende zur Reflexion über Softwarequalität anregen sollen. Hierzu wurde ein erster Prototyp entwickelt, der auf Basis von Quellcode und erkannten Soft- warequalitätsmängeln automatisch Lückentexte erzeugt. Die Untersuchung zeigt, dass dieser Ansatz grundsätzlich vielversprechend ist, gleichzeitig jedoch zahlreiche Herausforderungen bestehen. Die gewonnenen Erkenntnisse liefern erste Hinweise auf offene Forschungsfragen und verdeutlichen, dass für einen zuverlässigen Einsatz im Produktivbetrieb weitere Forschungs- und Entwicklungsarbeiten erforderlich sind.
17:00-17:20 Building Sustainable AI Competencies: Learning Concepts Grounded in Computational Thinking and Foundational Computer Science
Presenter: Eva-Maria Weiss and Franziska Paukner
Abstract: As digital and AI-driven tools have become ubiquitous, fostering competencies for reflective, critical, and responsible engagement is paramount for students in primary and lower secondary education. This paper presents a holistic educational approach based on AI literacy in fundamental computer science principles. The concept is grounded in building a deep understanding of informatics foundations and AI core concepts. A deeper understanding supports learners in developing Computational Thinking as well as accurate mental models of digital systems, and to cultivate epistemic vigilance. This enables students to engage critically and reflectively with AI applications, for example by understanding the central role of data and statistical learning processes in Machine Learning. By fostering a deeper understanding of the underlying mechanisms of AI systems, the approach aims to address common misconceptions, such as the attribution of human characteristics, emotions, or intentions to AI systems, and to reduce the risk of excessive trust in AI outputs and insufficiently critical evaluation of their results. By integrating unplugged activities with digital environments, we foster action competence, allowing learners to transition from passive consumers to autonomous, reflective, and empowered participants in a democratic digital society. Preliminary results from pilot implementations indicate that this approach appears to reduce misconceptions among students.
Call for Contributions
With this workshop, we aim to bring together CS education experts with diverse perspectives and facilitate dialogue between their intersecting subject areas to discuss their perspectives. The topics of the contributions include, but are not limited to:
- classroom applications and implementation of AI for CS education
- predictive and descriptive modeling for CS education
- adaptation, recommendations, and feedback in CS education
- learning analytics in CS.
- adapted competency requirements for AI in CS.
Submission Format: max. 4 pages (excl. references), LNI template. Please submit your contribution through the conference management system.
Organizers
Alina Deriyeva is a PhD student at Bielefeld University. Her research focuses on Intelligent Tutoring Systems for teaching programming, in particular the knowledge tracing methods and its applications. She was a part of the XLM (Explainable Learner Models) project, part of the KI:edu.nrw.
Sven Jacobs (M.Ed.) is a PhD student at the University of Siegen. His research focuses on formative feedback in CS education, particularly on leveraging GenAI to provide such feedback at scale and on understanding how students engage with it. He is also an active reviewer for several ACM Special Interest Group on Computer Science Education (SIGCSE) conferences.
Jesper Dannath is a PhD student at Bielefeld University. His research interests include next-step hints for programming tasks and intelligent tutoring systems. He was a part of the program committee for the Educational Data Mining Conferences in 2024 and 2025. He was also part of the organizing team of the SAIL workshop 'Fundamental limits of Large Language Models' in 2023.
Nadine Nicole Koch is a doctoral researcher at the University of Stuttgart. Her research focuses on optimizing feedback and gamification in intelligent tutoring systems. She is an active reviewer for the International Conference on Software Engineering Education and Training (CSEE&T).
Hendrik Fleischer (M.Ed.) is a PhD student at Leibniz University Hannover. His research in the field of chemistry education focuses on optimizing intelligent tutoring systems. The aim is to support learners effectively and adaptively in solving stoichiometry problems.
Prof. Benjamin Paaßen is junior professor for knowledge representation and machine learning at Bielefeld University, associated researcher at the Educational Technology Lab of the German Research Center for Artificial Intelligence (DFKI), Junior Fellow of the German CS Society, and member of the Young College of the Northrhine-Westphalian Academy of Sciences and Arts. Their research focus is explainable, interpretable, and domain-informed machine learning, especially for intelligent tutoring systems for CS education. They have chaired a wide range of special sessions and workshops at conferences, have been chair of the International Conference of Educational Data Mining (EDM2024), the 3rd TRR318 Conference, and two interdisciplinary spring schools.