Abstract
<jats:p>The article examines the expansion of the research agenda of psycholinguistics in the study of a communicative environment mediated by large language models (LLMs). It is argued that the widespread adoption of LLMs does not require a revision of the fundamental theoretical foundations of the discipline, but it does transform the empirical context in which speech activity is studied. The paper analyzes the specific features of user interaction with generative language systems and their implications for the analysis of interactive text production. It substantiates a shift in research focus from the isolated textual product to the dynamic process of its step-by-step formation under conditions of technological mediation. The study identifies several research directions that become particularly relevant in the context of human interaction with generative language systems, including: the perception of LLM-generated texts; calibration of trust and strategies for information verification; redistribution of cognitive effort; metacognitive control; dialogical dynamics; and the emergence of new communicative norms mediated by artificial intelligence. The article concludes that investigating this communicative configuration expands the empirical and methodological possibilities of psycholinguistics while preserving its theoretical continuity.</jats:p>