https://arxiv.publishscience.uz/id/eprint/1443 2026-06-04T05:00:06Z info:eu-repo/semantics/descriptiveMetadata https://arxiv.publishscience.uz/1443/ PRAGMATIC EQUIVALENCE IN AI TRANSLATION: CHALLENGES IN ENGLISH AND UZBEK DISCOURSE Karimova, Shaxnoza The rapid advancement of Artificial Intelligence (AI) and Neural Machine Translation (NMT) has significantly improved automated translation effectiveness. Nonetheless, achieving pragmatic equivalence remains challenging, especially when translating between linguistically and culturally diverse languages like English and Uzbek. This research examines the difficulties in maintaining pragmatic meaning in AI-based translations of English and Uzbek discourse. Using qualitative discourse analysis and comparative examples from popular AI translation tools, the study investigates politeness strategies, speech acts, implicature, cultural references, and context-dependent meanings. Findings reveal that while AI performs well with vocabulary and grammar, it often fails to capture intended meanings and cultural nuances. The paper emphasizes that integrating pragmatic understanding and cultural awareness is crucial for creating more accurate AI translation systems in the future. 2026-01-01 Article NonPeerReviewed application/pdf en https://arxiv.publishscience.uz/1443/1/fulltext.pdf Karimova, Shaxnoza (2026) PRAGMATIC EQUIVALENCE IN AI TRANSLATION: CHALLENGES IN ENGLISH AND UZBEK DISCOURSE. Scientific and innovative research in the social and humanitarian sphere, 3 (5). pp. 640-644. ISSN 3060-5059 OJS:1678 OJS:1678 info:eu-repo/semantics/objectFile info:eu-repo/semantics/humanStartPage