下游市场中的人工智能与竞争动态

OECD Roundtables on Competition Policy Papers, No. 331 Artificial intelligence and competitive Disclaimers This work is published under the responsibility of the Secretary-General of the OECD. The opinionsexpressed and arguments employed herein do not necessarily reflect the official views of the Member This document, as well as any data and map included herein, are without prejudice to the status of orsovereignty over any territory, to the delimitation of international frontiers and boundaries and to the Photo credits:©shironosov/ iStock-Getty Images Plus © OECD 2025. Attribution4.0 International (CCBY4.0) This work is made available under the Creative CommonsAttribution4.0 International licence. By using this work, you accept to be bound by the terms of this licence(https://creativecommons.org/licenses/by/4.0/). Attribution–you must cite the work. Adaptations–you must cite the original work and add the following text:This is an adaptation of an original work by the OECD. The opinions expressed and arguments Third-party material–the licence does not apply to third-party material in the work. If using such material, you are responsible for obtaining permission from the third party Abstract This paperexamines how the adoption of artificial intelligence (AI),particularly generative and agentic systems, is reshaping competition indownstream markets.It explores mechanisms through which AI may lowerbarriers to entry, substitute for labour, reduce minimum efficient scale, and The paper analyseshow AI affects market structure andmay shapefirmbehaviour, finding that itscompetitive impact is highly context-dependent,shaped by sectoral exposureto AI use, firmsize andcapabilities, andaccess to enabling inputs.Itconcludes by discussing enforcement, Keywords:Agentic AI,Algorithmic collusion,Artificial Intelligence,Barriers to entry,Competitionenforcement in AI,Data access,Efficiency and productivity gains, Generative AI,Market contestability, JEL codes: L40, L86, D43, D82, O33, K21 Thispaper is part of the series“OECD Roundtables on Competition Policy Papers”,https://doi.org/10.1787/20758677. Acknowledgements This note was prepared byAnia Thiemannof the OECDCompetition Division. Helpful comments andreview were provided by Ori Schwartz, Antonio Capobianco,Richard May and Connor Hogg, also of theOECD Competition Division.This paper is part of the OECD Horizontal Project on ‘Thriving with AI:Empowering Economies, Societies and Citizens.’It has also benefited from useful comments fromcolleagues in the OECD Directorate forScience Technology and Innovation, particularlyJulia Carroat the Table of contents Disclaimers 3 Abstract Acknowledgements Executive summary AI as a general-purpose technologyMechanisms supporting competition 1 Introduction 1.1. Key concepts used 2 Theimpact of AI adoption on market dynamics 2.1. What are the mechanics by which AI can foster competition?2.2. Limits to the usefulness of AI systems2.3. Sector exposure and uptake of AI vary significantly 3 AI-related competition concerns in downstream markets 3.1. Competition enforcement concerns3.2. An emerging concern: the attribution of liability 4 Conclusion and future directions References39 50 Notes BOXES Box1. AI and knowledge worker substitutionBox2. Use of AI in the manufacturing sectorBox3. AI-driven credit scoring can boost alternative financeBox4. Experimental evidence on the productivityeffects of generative AIBox5. Generative AI for agricultural support in MalawiBox6. AI uptake and diffusion in downstream sectors in India Executivesummary This paper examines how the adoption of artificial intelligence (AI), particularly generative and agenticsystems, is reshaping competition in downstream markets. While much of the policy debate has focusedon competition in AI infrastructure and foundation model development, this paper considers how AI is used The analysis suggests that AI may support competition in downstream markets by lowering barriers toentry, reducingminimum efficient scale, and enabling product differentiation and innovation. At the sametime, it identifies emerging risks related to data access, model restrictiveness, vertical integration and AI as ageneral-purposetechnology Generative AI (GenAI) exhibits characteristics of a general-purpose technology, including pervasiveness,continuous improvement and innovation-spawning potential. Estimates of productivity gains vary, withstudies suggesting annual increases in total factor productivity (TFP) ranging from 0.07 to 1.3 percentage Mechanismssupportingcompetition AI adoption may foster competition through several channels: Labour substitution and augmentation:GenAI systems can automate cognitive tasks,particularlyroutine or repetitive tasks,lowering the skills threshold for market participation. Field experimentsalsosuggest significant productivity gains, particularly for less experienced workerswhen they use AI to helpperform higher-skilled tasks.