Situated appropriation of ChatGPT among students: An experience sampling study of recurring use contexts

Generative AI is becoming embedded in everyday communication environments, yet research still often treats its use as a stable adoption outcome rather than as a situated communicative practice. Drawing on Human-Machine Communication and appropriation research, this study investigates students’ ChatGPT use as situated appropriation in everyday life. Empirically, we combine a two-wave experience sampling design with multilevel latent class analysis. The study includes 59 university students and 522 reported ChatGPT-use episodes. We identify four recurring generative-AI communication situations (home-based desktop task use, home-based mobile informal use, workplace desktop use, and public mobile use) as well as three user repertoires that capture how individuals distribute their use across these situations. These findings provide a situational map of students’ ChatGPT appropriation and show that communicative AI use is organized through recurring context configurations and person-level repertoires rather than a single, unitary pattern of use. The study highlights the value of treating situations as a key unit of analysis for understanding how communicative AI becomes embedded in everyday routines, institutional settings, and evolving expectations toward machine-generated contributions. It also demonstrates the usefulness of combining experience sampling with configuration modeling to study human–AI communication in context.