The arrival of commercially available AI systems has prompted users’ exploration. As the scope of the tasks undertaken and supported by these AI artifacts increases, different forms of use emerge, driven by the user’s processes of trial and discovery of relevant tasks and use cases. While IS research has extensively examined exploratory use, we argue that the learning and discovery processes for general purpose AI systems (such as generative AI chatbots) may be different. Thus, this study addresses the research question: “What are the different types of exploratory AI use and how do they evolve over time?” We present the theoretical foundations of exploratory use of AI, propose a typology of exploratory use of AI, and propose a methodology to test the validity of the typology and assess how exploratory use changes over time.