The integration of chatbots in higher education has emerged as a transformative tool to increase student engagement, streamline administrative processes and support personalised learning. This article examines the current uses, benefits and challenges of chatbot technology in academia, based on an empirical study of unique sample of 408 students at a pedagogical university, people who will shape the future education of others. The findings indicate that students are most likely to use ChatGPT, and to use it mainly in education-related areas, and to a lesser extent in areas of everyday life. In this respect, non-pedagogy students were more likely than pedagogy students to generate ready-made answers during exams. Despite students claiming to have a fairly good knowledge of chatbots, only a quarter had read articles or research on them. The study also explored ethical issues related to students’ use of chatbots. Few respondents believe that their teachers or university have rules or guidelines for the responsible use of chatbots. Non-pedagogy students and pedagogy students with teaching qualifications rate this issue the lowest. Almost half of the respondents feel privacy and security concerns when using a chatbot. Seven out of ten believe that chatbots can be used to manipulate information or mislead students. Students also believe that chatbots could increase inequalities in education. Through a synthesis of existing literature and practical implementations, this paper provides insights for teachers, administrators and developers who want to use chatbot technology to foster innovation in university education.