Daniel O'Sullivan, Davide Ravasi
Debates about how generative artificial intelligence (GenAI) will impact the practice of qualitative methods have largely focused on how GenAI may be used to collect and analyze data, as well as the ethical implications of doing so. In this paper, we shift attention to a less-examined but vital aspect of qualitative research: the interpretation – or theorization – of empirical patterns. We argue that GenAI’s core capabilities – autoregressive generation, self-attention, and latent knowledge – make it particularly well-suited to augment qualitative theorization. Based on an understanding of qualitative theorization as disciplined imagination, we elucidate how GenAI can support this process by expanding and enriching imagination and broadening and scaffolding discipline. We offer practical guidance on how to prompt GenAI to realize these benefits, whilst also identifying potential pitfalls and how to avoid them. Ultimately, we contend that, although GenAI is no substitute for human interpretation, it represents a powerful tool for extending and structuring the theorization process.