HUMAN-AI INTERACTION IN MULTILINGUAL PROFESSIONAL

The article offers a comprehensive pragmatic analysis of human – artificial intelligence (AI) interaction in multilingual professional teams operating in digital and remote work environments. Drawing on Speech Act Theory (Austin, Searle), Grice’s Cooperative Principle, and Goffman’s face-work theory as reinterpreted for digital discourse, the study examines how AI-mediated communication reshapes pragmatic patterns across language and cultural boundaries. The empirical base consists of 240 anonymized interaction logs from international remote teams that actively used AI assistants – including ChatGPT, Microsoft Copilot, and DeepL Write – over the period 2021-2024. Participating teams comprised members from at least three distinct linguistic backgrounds (Anglophone, Hispanophone, and Germanophone regions). Four dominant pragmatic strategies are identified in the corpus: reformulation, hedging, metalinguistic commentary, and code-switching facilitation. The article argues that contemporary AI tools function not merely as passive translation instruments but as active pragmatic mediators that partially compensate for intercultural miscommunication while simultaneously introducing new forms of pragmatic asymmetry and communicative impoverishment. Algorithmic pragmatic failures – particularly the systematic misreading of irony and culturally specific connotations – are identified and theorised as a distinctive feature of AI-mediated discourse. Implications for translation studies, intercultural communication theory, applied linguistics, and corporate communication practice are discussed.