Multilingual health communication with Generative AI: Students as translators/ adaptors
The study presents a compelling case for rethinking how multilingual health communication is produced by exploring the concept of “citizen adaptors” – non-professional bilingual individuals who collaborate with generative AI to translate health messages.
Through workshops with 22 university students in Australia and Hong Kong who adapted neurofibromatosis information into Chinese Mandarin, Cantonese, Indonesian, and Russian, the researchers found that effective multilingual health adaptation requires three key capacities: reciprocity (the ability to teach and learn from AI through feedback loops), biliteracy skills (not just speaking but reading and writing proficiency in multiple languages), and affective attachments to the intended audience (emotional connection that motivates quality work).
Drawing on feminist theorist Donna Haraway’s concept of “response-ability” – the capacity to respond ethically to complex challenges – the study challenges traditional binaries that position professional translation as superior to adaptation work done by community members with AI assistance.
While the research identified important limitations, including difficulties with technical terminology and the time-intensive nature of quality adaptation work, it suggests that citizen adaptors could provide a viable, cost-effective alternative for not-for-profit organizations that cannot afford professional translation services, particularly during health crises when rapid multilingual communication can be life-saving.
Read more here.
H. Heinrichs, D., Camit, M., & Tsao, J. (2025). Response-able, collaborative adaptations of multilingual health messaging: a case study from Australia and Hong Kong, SAR. Critical Public Health, 35(1). https://doi.org/10.1080/09581596.2025.2555211