In our study, ChatGPT was prompted with the theoretical meaning of the term using the researcher’s understanding of the concept. The issue you highlight was actually documented in our paper: an initial prompt we used was misinterpreted by ChatGPT as a more general definition than the theoretical definition, necessitating that the researcher revise the prompt. The implicit meaning we found ChatGPT could identify was seen in multiple places, where it interpreted what the person said, connected it to other previous answers, and then saw whether it could reflect something even if it wasn’t explicitly discussed. We agree that the disagreements should definitely be retained, as a way for the researcher to further interrogate the data. This can demonstrate an audit trail of decision-making in the research process. Incorporating the participant in the research process, as we did, further demonstrates a participatory approach to interrogate the GenAI interpretations, which could prompt the participant to challenge GenAI’s analysis or to adapt their own understanding based on new information.