Abstract
<jats:p>This article examines the problem of equivalence in digital translation based on English and Uzbek language materials. Today, automatic translation systems such as Google Translate, DeepL, and Yandex Translate significantly accelerate the translation process; however, they do not always ensure full lexical, semantic, grammatical, stylistic, and linguocultural equivalence. Particular difficulties arise in translating phraseological units, national and cultural realia, polysemous words, context-dependent expressions, and literary-stylistic devices. As a result, digital translation may lead to loss of meaning, incorrect equivalent selection, and stylistic inconsistency. The article analyzes the main types of equivalence, their linguistic characteristics, and the key problems observed in English-Uzbek and Uzbek-English translation. Special attention is paid to comparing machine translation and human translation results, as well as to the importance of considering contextual, pragmatic, and cultural factors in improving the quality of digital translation.</jats:p>