Purpose: This study aims to investigate intercultural communication failures in AI-generated translations between English and Uzbek, focusing on the preservation of pragmatic meaning, cultural nuances, and communicative intentions.
Methodology: The research employs comparative, pragmatic, and discourse analysis methods to examine AI-translated idiomatic expressions, humor, politeness strategies, and culture-specific units. AI systems analyzed include ChatGPT, Google Translate, and DeepL.
Results: The study found that AI-powered translation systems often rely on literal and structural translation strategies, which frequently lead to semantic distortion and pragmatic inaccuracies. As a result, intercultural misunderstandings and communication failures occur, especially between linguistically and culturally distant languages.
Conclusions: Integrating intercultural competence, cultural mediation, and context-sensitive mechanisms into AI translation models is crucial to enhance pragmatic equivalence and sociocultural adaptation in multilingual communication.
Limitations: This study primarily analyzes AI translations in English-Uzbek contexts and relies on qualitative methods; broader empirical testing across other languages and larger datasets is needed.
Contributions: The study contributes to modern translation studies by elucidating the relationship between artificial intelligence, intercultural communication, and translation pragmatics, offering insights for the development of culturally aware and context-sensitive AI translation technologies.