Understanding AI Agents—A Data-Driven Literature Review

This paper presents a systematic, data-driven literature review of research on Artificial Intelligence (AI) agents based on the top 100 Google Scholar publications related to the search terms “AI agent” and “AI agents”. The rapid advancement of AI agents, driven in particular by recent progress in Large Language Models, has resulted in a diverse and fragmented research landscape that lacks comprehensive quantitative overviews. To address this gap, we implement and apply a fully automated, AI-driven analysis pipeline to the domain of AI agents. The collected publications are processed using a Large Language Model accessed via a Python-based Application Programming Interface (API), enabling an automated analysis of the literature without manual categorization. Based on this approach, the publications are grouped into data-driven thematic clusters reflecting dominant research perspectives in the field. Specifically, the identified clusters comprise “Architecture & Frameworks”, “Multi-Agent Systems”, “Applications”, “Safety” and “Ethics, Accountability & Governance”. By synthesizing the literature in a structured and automated manner, this work provides a consolidated overview of central research patterns, identifies key operational and structural challenges and highlights fragmentation across AI agent research. The findings support a more systematic understanding of AI agents and provide a foundation for future research on robust, scalable and trustworthy AI agent systems.