Исследования ИИ-агентов в экономике и компьютерных науках: тематические направления, динамика и перспективы

Бизнес-информатика. 2026. Т. 20. № 3. С. 47–68.

Naidenova I. N., Parshakov P., Smirnov A., Shakina E., Shenkman E.

This article presents a systematized review of academic publications on autonomous AI agents, based on the analysis of 1,634 works from international peer-reviewed journals indexed in the Scopus database. Using thematic modeling with contextualized embeddings via BERTopic, the study identifies key research directions on AI agents within computer science as well as their applications in the social sciences. Thirteen major thematic areas are distinguished and grouped into the following broader categories: interaction and communication of AI agents, applied use of AI agents, ethics and trust in AI agents, and the learnability and adaptability of AI agents. The review highlights a growing interest in the study of human-AI agent interaction and reveals a thematic evolution from formal planning models toward the integration of large language models and self-learning systems. The analysis also points to the dominance of U.S.-based organizations in the volume of published research, alongside the rapid development of this field in China, with all other countries significantly lagging behind. The article concludes by outlining promising research directions, including the coordination of heterogeneous agents under uncertainty and the assessment of the socio-technical implications of their deployment.

Keywords: социальные науки, компьютерные науки, тематическое моделирование, систематизированный обзор литературы, ИИ-агенты, BERTopic, развитие направлений исследований, страновые исследовательские тренды