ARTICLE
4 August 2026

The Potential Impact Of Agentic AI On Competition And Merger Review

N
NERA

Contributor

Since 1961, NERA has provided unparalleled guidance on the most important market, legal, and regulatory questions of the day. Our work has shaped industries and policy around the world. Our field-leading experts and deep experience allow us to provide rigorous analysis, reliable expert testimony, and data-powered policy recommendations for the world’s leading law firms and corporations as well as regulators and governments. Our experience, integrity, and economic ingenuity mean you can depend on us in the face of your biggest economic and financial challenges.
As agentic AI systems gain the ability to execute complex tasks with minimal human intervention, competition authorities face new challenges in understanding how these technologies reshape competitive dynamics. This analysis examines the dual nature of AI agents—their potential to both intensify competition through efficiency gains and raise concerns about algorithmic coordination and market concentration.
United States Antitrust/Competition Law
NERA are most popular:
  • within Intellectual Property, Finance and Banking and Environment topic(s)
  • with readers working within the Technology industries

As agentic AI becomes capable of carrying out increasingly complex tasks with limited human oversight, competition authorities may need to consider how these technologies could influence competitive interactions and dynamics. In “The Potential Impact of Agentic AI on Competition and Merger Review” for the July issue of Competition Policy International, Directors Cristián Hernández and Gabriella Monahova explore these emerging issues and their implications for the assessment of horizontal mergers. 

Dr. Hernández and Dr. Monahova explain that agentic AI has the potential to both strengthen and weaken competition, depending on how it is adopted and integrated into existing operations and market interactions . While AI agents may reduce transaction costs, improve efficiency, and intensify competition in some markets, they may also raise concerns related to personalization, tacit coordination, and the entrenchment of market power. The article emphasizes that these effects should be assessed on a case-by-case basis. 

The authors conclude that, although the traditional framework for merger review remains applicable, assessments may increasingly require evidence on firms’ AI capabilities and how they are deployed. The article concludes by offering a practical framework for evaluating how emerging AI technologies may shape competition and merger outcomes as agentic AI continues to evolve.

Key Takeaways

  • Firms’ and consumers’ use of agentic AI could increase competition through cost reductions and quality enhancements, but it could also raise competitive concerns, such as algorithmic collusion risk.

  • The competitive effects of agentic AI will depend on the specific application, including whether adoption and integration occur evenly across firms and consumers or reinforce existing asymmetries. 

  • While the traditional framework for horizontal merger review remains applicable, the assessment of competitive effects will require additional evidence on firms’ AI capabilities and how they are deployed.

How NERA Can Help

NERA helps clients solve complex, data-intensive business, legal, regulatory, and economic questions by combining expert judgment with the latest data engineering, cloud analytics, machine learning, and responsible AI. Our researchers, analysts, data scientists, and expert economists turn large and sensitive datasets into defensible insight while operating within a mature security, confidentiality, and privacy framework specifically designed for high-stakes client work.

Our antitrust and competition teams analyze how AI capabilities affect the antitrust risk of a proposed merger, evaluate whether use of agentic AI is likely to enhance or weaken competition, and prepare defensible economic analyses of AI-related issues in litigation. 

We accomplish this using a sophisticated technology model built for matters in which data volume, analytical rigor, speed, and confidentiality all matter. We bring together economists, consultants, and data scientists to design workflows that are scalable, auditable, and tailored to each client’s needs. 

  • Client-specific analytics environments: On-demand Databricks workspaces can be provisioned for individual client teams, safeguarding project work and reducing the risk of inappropriate data visibility across matters. You can be confident your data stays with your team. 

  • Flexible, secure data ingestion: We can ingest and exchange data through solutions such as Delta Sharing, DataSync, and AWS Transfer. We work with clients to align with their technical, legal, and operational requirements. 

  • Modern lakehouse architecture: NERA uses a lakehouse approach with medallion-style architecture to organize data pipelines from raw ingestion through curated, analysis-ready datasets, supporting reproducibility, governance, and efficiency.  

  • Responsible AI acceleration: With client approval, teams can use enterprise AI tools such as ChatGPT Enterprise, GitHub Copilot, and Databricks Genie to accelerate coding, analysis, documentation, and project delivery while keeping experts in control. Our enterprise agreements with approved AI vendors include contractual protections designed to safeguard confidential data. 

  • Human-validated insight: AI is treated as a productivity multiplier, not a substitute for professional judgment. NERA experts validate and challenge AI-enabled outputs before those outputs inform client work. 

  • Dedicated data science expertise: A specialized data science team supports case teams with advanced programming, data engineering, machine learning, automation, reusable firm-wide tools, and technical subject matter expertise. 

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.

[View Source]

Mondaq uses cookies on this website. By using our website you agree to our use of cookies as set out in our Privacy Policy.

Learn More