Session T-17 - Topics: Responsible Generative AI Use; Student Perceptions of AI Marking

Generative AI is rapidly reshaping programming education, but instructors still lack practical frameworks for deciding how to integrate it responsibly. This proposal introduces a design-space framework organized around two pedagogically decisive dimensions: when AI enters the learning process, and how much cognitive work it is allowed to perform. Drawing on an ongoing synthesis of recent empirical studies, we outline twelve instructional configurations and discusse how different design space positions imply different risks, governance needs, and opportunities for meaningful learning.

Institutions are increasingly exploring AI to mark open ended questions, yet little is known about how students perceive this shift. Concerns include fairness, motivation, and academic integrity, particularly in large online courses where consistent marking is challenging. This presentation will report on a study examining student perceptions of AI based marking through surveys administered at two points in the term, highlighting perceived benefits, concerns, and students’ willingness to accept AI supported assessment.