The authors argue that at present, a new scientific object, a discourse with the key term artificial intelligence (AI), has formed in the language of science. The state of the discourse is characterized as extremely debatable. From a substantive point of view, the articles have the object of AI technology, and the subject is the problems of the emergence of AI systems in various branches of human activity. The formation of AI discourse is at the initial stage, as evidenced by the elements of linguocreativity in the texts of institutional discourse, characterized by the lack of imagery, personalities and terminology. The authors emphasize the novelty of this phenomenon and the comprehensiveness of AI technologies with different words and word combinations with temporal and spatial semantics. In the discourse lexical means create the effect of speed, pressure, quantitative and qualitative growth of new AI systems. The key words of the new discourse are advantage, benefit, trust, challenges. The discourse is characterized by metaphorical language and a large number of personifications. From axiological positions the value of AI is not questioned. From the point of view of determining the author’s goal-setting, the purpose of the text in the discourse, new goals are affirmed, namely the justification of the opportunities and limits of the application of AI. The communicative strategy of the authors of the articles is aimed at creating the effect of readers’ involvement, synergy in achieving the goals. For this purpose, a variety of rhetorical means and techniques are used (personal pronouns, personal forms of verbs, rhetorical questions, oxymoron, phraseological expressions, winged words, epithets, etc.). At this stage of the existence of scientific discourse with the key word artificial intelligence is found similar to the artistic discourse, as there is the creation of a new scientific object, looking for approaches to its linguistic positioning, forms of putting forward the scientific result.