Generative AI and chatbots in university education

The public release of ChatGPT-3.5 in November 2022 marked a turning point for higher education. Although not designed for educational use, generative AI chatbots quickly entered university classrooms, leading to both enthusiasm and concern. This chapter presents a systematic review of peer-reviewed empirical research published between 2018 and January 2025 that examines the development and use of generative AI and large language models within the humanities and social sciences in higher education. Drawing on our background in anthropology and social studies of science and technology, we focus on how generative AI reshapes pedagogical practices and learning processes. From an initial pool of 16,403 studies, 159 met our inclusion criteria. Our analysis identifies four main themes: (1) new perspectives on learning theory and pedagogy; (2) the role of teachers and t-bots, including automated assessment; (3) student engagement with generative AI; and (4) debates about AI literacy in relation to digital and academic literacies. We find that most existing studies emphasize technical aspects or quantitative outcomes rather than teachers’ and students’ everyday practices through which AI is used in education. The chapter concludes that future research needs to engage more deeply with practice-based and qualitative approaches to understand how generative AI transforms higher education as a socio-cultural and pedagogical field.