Text generation as an object of linguistic research

Th e article is an overview of research in the fi eld of text generation by artifi cial intelligence. Th e relevance of the topic is due to the wide application of neural networks in media, education and other spheres, which requires a systematic study of generated texts. Th e aim of the paper is to characterise the main directions of research of generated texts, including technological, ethical and linguistic directions. Th e methods used include descriptive, comparative and contextual analyses. As a result of the study of existing research, the most signifi cant aspects are identifi ed: training of language models, natural language processing, legal regulation of authorship of texts generated by artifi cial intelligence and elimination of cultural distortions. Special att ention is paid to linguistic aspects, as well as the possibilities of practical application of generated texts. Th e scientifi c novelty consists in a comprehensive approach to the study of the phenomenon of text generation, combining linguistic, technological and ethical perspectives. Th e practical signifi cance of the work is to identify directions for further research, for linguistic monitoring of one of the abilities of neural networks in the form of text generation.