Contact Language and ChatGPT How does artificial intelligence interpret the use of Hungarian in Vojvodina?

Abstract

For decades, Hungarian linguists have researched the phenomena of Hungarian language use in Serbia (Vojvodina). However, with the spread of artificial intelligence, a knowledge gap has emerged, which was identified and possibly addressed by the online linguistic case study assessing the current possibilities of interpreting Hungarian language use in Vojvodina by ChatGPT (Balázs –Beke–Szűts, 2025), published in the July issue of Magyar Nyelvőr. This research report is an edited and condensed English version of the cited paper. The English language of the main text enables a generalised, objective presentation of the conducted empirical study and the drawing of scientific conclusions. For the sake of followability, the present English text presents the user prompts, which constitute the corpus of the empirical research, containing lexical elements and other characteristics from the Serbian language, in an unedited form, and then their colloquial interpretation by artificial intelligence.