Paper Type
Short
Paper Number
PACIS2026-1512
Description
Generative Artificial Intelligence (GenAI) is challenging traditional assumptions about authorship, originality, and the demonstration of competence in higher education. In Information Systems (IS) education, AI-assisted work is increasingly embedded in both learning processes and professional practice. Institutional responses have largely focused on detection tools and academic integrity policies; however, these approaches remain reactive and technologically unstable. This paper argues that the challenge is architectural rather than behavioural. Drawing on programmatic assessment literature and socio-technical systems theory, the study reconceptualises assessment as a system of evidence production, interpretation, and judgment. The paper proposes a three-layer framework for AI-resilient assessment governance comprising: assessment design archetypes that represent different modes of capability demonstration, an AI vulnerability exposure framework that analyses structural susceptibility to AI-supported substitution, and a composite capability assurance model that integrates multiple forms of evidence. The framework contributes a conceptual foundation for redesigning assessment in AI-augmented IS education contexts.
Recommended Citation
Madugoda Gunaratnege, Senali Lecturer, "Socio-Technical Framework for AI-Resilient Assessment in Information Systems Education" (2026). PACIS 2026 Proceedings. 6.
https://aisel.aisnet.org/pacis2026/is_education/is_education/6
Socio-Technical Framework for AI-Resilient Assessment in Information Systems Education
Generative Artificial Intelligence (GenAI) is challenging traditional assumptions about authorship, originality, and the demonstration of competence in higher education. In Information Systems (IS) education, AI-assisted work is increasingly embedded in both learning processes and professional practice. Institutional responses have largely focused on detection tools and academic integrity policies; however, these approaches remain reactive and technologically unstable. This paper argues that the challenge is architectural rather than behavioural. Drawing on programmatic assessment literature and socio-technical systems theory, the study reconceptualises assessment as a system of evidence production, interpretation, and judgment. The paper proposes a three-layer framework for AI-resilient assessment governance comprising: assessment design archetypes that represent different modes of capability demonstration, an AI vulnerability exposure framework that analyses structural susceptibility to AI-supported substitution, and a composite capability assurance model that integrates multiple forms of evidence. The framework contributes a conceptual foundation for redesigning assessment in AI-augmented IS education contexts.
Comments
04-DigitalLearning