Generative AI presents significant challenges for Computer Science assessment, raising questions about validity, authenticity, and inclusivity across diverse assessment formats. This RIPPA brings together computing educators to co-create and apply a practical framework for evaluating assessment in relation to generative AI, generating a shared evidence base and practical guidance to support informed and inclusive assessment design in AI-rich education.
The rapid development of generative AI tools has introduced chal lenges for assessment in Computer Science in Higher Education, extending longstanding concerns around academic integrity and the limitations of detection-based approaches 3, 6, 8 . More recently, computing education research has examined the implications of generative AI for programming education, highlighting both its educational potential and the risks it poses for assessment validity 1, 4 . UK sector bodies have similarly called for assessment to adapt while maintaining standards and supporting learning 2, 5, 7 .
Existing approaches largely provide general principles for assessment reform or frameworks for specifying permitted levels of AI use. There is still a gap in relation to a Computer Science-specific framework that can be used to evaluate how generative AI affects the breadth of assessment used across the discipline, from programming coursework to written and oral examinations, presentations and practical tasks. In the absence of such a framework, reactive responses, including increased reliance on invigilated examinations, risk narrowing assessment and undermining inclusive practice.
This RIPPA addresses the gap through a collaborative, multiinstitutional process to co-create and refine a framework for evalu ating assessment in response to generative AI. It considers contexts in which AI may be restricted, regulated, or incorporated, and will produce a shared evidence base and practical tools to support assessment evaluation and redesign.
The project will be guided by the following research questions:
The project will use a participatory, iterative co-design approach organised in three phases. Participants will help to develop the framework, apply it within their own contexts, and contribute to shared cross-institutional analysis.
The RIPPA will be led by a team of computing education researchers and practitioners from eight UK HEIs, with expertise in assessment, pedagogy, generative AI, and curriculum design. The Project Team will coordinate workshops, support participants in applying and refining the framework within their own contexts, and lead the synthesis, evaluation and dissemination of findings.
A wider group of computing educators will be recruited at UKICER 2026 and involved throughout the project. By applying and refining the framework through their own assessment practice, participants will contribute to its refinement across diverse institutional contexts and assessment formats, while producing outputs that are directly applicable to their own teaching practice. Through participation, individuals will become members of a Community of Practice dedicated to designing and refining Computer Science assessment in response to generative AI. Through workshops, peer review, and cross-institutional sharing of assessment redesigns, participants will apply a structured evaluation framework, receive constructive feedback on their own assessment practices, and learn from colleagues across diverse institutional contexts.
The project will produce a co-created, practice-informed framework for evaluating the impact of generative AI on Computer Science assessment, alongside a set of practical assessment design principles and a collection of cross-institutional case studies illustrating their application in different contexts. Together, these outputs will provide both a shared evidence base and practical guidance to support educators in reviewing and adapting assessment in response to AI. The project will lead to a co-authored publication submitted to UKICER 2027. The framework, case studies, and supporting materials will also be shared openly with the wider computing education community to support ongoing discussion and development of assessment practice in Computer Science education.