Challenges of communicative AI in education

Education has been a testbed for new developments in Artificial Intelligence since its beginnings in the 1950s. After years of slower progress with Intelligent tutoring and adaptive learning systems, the advent of communicative AI such as ChatGPT revitalised the concepts of automated feedback to support individualised learning. Education research has shown that feedback can improve learning, particularly formative feedback. Based on a prototype of an automated feedback system for multimodal learning results in higher education settings, the presentation will address challenges of designing and implementing these systems. These range from questions about biases in the data and the models to regulatory aspects (about privacy and copyright) and its organisational embedding. And with the automation of communication in teaching and learning settings, the roles of teachers, learners and technologies will be re-assigned.

Andreas Breiter is a full professor for information management and educational technologies at the University of Bremen (Germany) in the Department for Mathematics and Informatics. He is Scientific Director of the Institute for Information Management Bremen ( and leads the lab on socio-technical systems and critical data studies at the Centre for Media, Communication and Information Research (ZeMKI, His research work centers around processes of datafication and organizational change within the educational system, combining technological design with critical data studies.