The paper explores speech-to-speech interpretation systems (SISs) application to service multilingual communication in humanitarian contexts related to forced migration. The study compares various systems capacity to process non-native speakers’ speech to map major challenges for the above systems use within the mentioned settings. The research introduces interim results of a pilot study in terms of research sample (non-native speakers and interpreters), the selected language pair, and the comparative list of recommender platforms. The research includes the relevant literature study, comparative analysis of SISs outputs regarding the interpretation of non-native speakers’ accented speech, when interpreted from English into Russian, interpreters’ surveys on the above analysis results. The technology under study included Google Translator, Microsoft Translator, and Yandex. The pool of research participants included refugees from different countries and professional interpreters. The research rested on comparative qualitative multidimensional analysis, integrated content-based selection of academic sources and their theoretical analysis, descriptive empirical analysis of language errors by SISs, interpreters’ survey through open-ended questionnaire, factor, cluster, and content analysis to process their replies. The results map those language and communicative context features that should be considered for digital interpreting systems further tuning in terms of multilingual instrumentation, set forth the tasks to develop the relevant methodology for further studies, customize the technology to specific communicative settings and to train specialists to use it in socially critical contexts. © 2022, Springer Nature Switzerland AG.