This paper identifies the most frequently cited features in existing definitions of agentic AI and AI agents, examines how these features are described across sources, and maps them to the key elements of the OECD definition of an AI system. By highlighting both shared traits and differences, the paper aims to support clearer conceptual understanding and inform future research and policymaking. It also provides descriptive data on recent trends in the uptake of AI agents and agentic AI.
The agentic AI landscape and its conceptual foundations
Working paper
Share
Facebook
Twitter
LinkedIn
Abstract
In the same series
-
Working paper
Evidence from selected countries and the European Union
7 May 202658 Pages -
Working paper
Global linkages and the cross‑country distribution of the gains from AI
18 March 202679 Pages -
Working paper
International insights and policy considerations for Italy
11 December 2025100 Pages -
8 December 202543 Pages
-
1 December 202530 Pages
Related publications
-
Working paper
Evidence from selected countries and the European Union
7 May 202658 Pages -
Working paper
Global linkages and the cross‑country distribution of the gains from AI
18 March 202679 Pages -
19 February 202669 Pages