Trust and Usage of LLMs in the BCMS Language Sphere

The project investigates how speakers of Bosnian, Croatian, Montenegrin, and Serbian (BCMS) languages interact with and trust large language models like ChatGPT, compared to their usage in English. We will survey a large number of participants across four BCMS countries to understand language preferences, trust perceptions, and usage patterns. The findings will contribute to our understanding of linguistic inclusivity in AI development and help inform future digital policy in the Western Balkans.

I am working ong this project in a collaboration with Marina Budić, PhD.

The project is being funded by The Open Society Foundation, Western Balkans.