ГЕНЕРАТИВНЫЙ ИСКУССТВЕННЫЙ ИНТЕЛЛЕКТ В РЕКЛАМНОМ ДИСКУРСЕ: ЯЗЫКОВЫЕ СДВИГИ И ТРАНСФОРМАЦИЯ ГОЛОСА БРЕНДА

Abstract

This article examines how generative reformulation changes the linguistic organization of advertising messages and the realization of their persuasive function. The study is relevant because generative models are being rapidly integrated into advertising copywriting, while large-scale content production may lead to stylistic homogenization. The aim is to identify differences between original advertising posts and AI-generated versions across three groups of indicators: imagery and metaphorization, persuasion strategies, and markers of brand voice. The dataset comprises 40 English-language messages published by official brand accounts and 40 comparable generated versions that preserve the original factual content and communicative purpose. The study uses content analysis, manual coding, and paired statistical comparison. The AI-generated versions show a higher concentration of sensory-evaluative vocabulary, more frequent sequencing of product benefits, and greater syntactic elaboration. At the same time, they contain fewer direct calls to action, urgency markers, sentence fragments, colloquial expressions, emojis, and instances of expressive punctuation. The findings indicate a shift from direct prompting toward a more extended presentation of benefits. The article concludes that generative AI facilitates content scaling but may weaken distinctive features of brand voice when stylistic parameters are not specified explicitly.