Fair and transparent dialogic assessment through AI in Higher Education - CAISA Fellowship

This CAISA Fellowship (The National Center for AI in Society) will investigate how AI can enhance fairness and transparency in oral assessment practices in Higher Education. Worldwide, oral assessment is facing a revitalization due to generative AI’s ability to produce text. Even though oral exams are often presented as objective measures, they are still conditioned by subjective beliefs and values.

The fellowship examines how AI can create meaningful traces of dialogic assessment. Rather than replacing human judgment, these AI tools will be used to augment assessors’ decision-making and make the assessment process more transparent. The fellowship will synthesize emerging practices on AI-supported dialogic assessment, and explore a prototype developed in collaboration with USN & OsloMet.