This article provides an in-depth analysis of the types of errors that occur in artificial intelligence-based translation systems between English and Uzbek. The study systematically classifies lexical, grammatical, semantic, pragmatic, and cultural errors and explains their causes from linguistic and cognitive perspectives. The paper also examines the capabilities and limitations of neural machine translation systems. The findings indicate that although AI translation systems have approached formal accuracy, they still face challenges in ensuring pragmatic and cultural equivalence.
Shakhnoza Valievna Karimova (Mon,) studied this question.