Algorithmic Price Coordination in Competition Law

February 2026 | Başaran Law Firm

I. Introduction

Can two competing companies raise their prices in a coordinated manner without a single phone call, email, or meeting in the same room? The answer is yes — if both are using the same algorithm. This phenomenon, which competition authorities have termed "collusion by code," is pushing the boundaries of traditional competition law.

II. Why Does Algorithmic Pricing Matter for Competition Law?

A pricing algorithm is not inherently harmful. An e-commerce platform updating its prices within milliseconds can generate both efficiency gains and consumer benefits. The problem emerges when algorithms interact with one another. Systems that update their own behaviour by learning from market responses can, over time, learn to converge at price levels above those that would arise under normal market conditions — all without any person agreeing with any other. The algorithm learns on its own.

A 2025 study by Wharton School researchers Hangcheng Zhao and Ron Berman adds an important nuance: in markets where consumers can compare a greater number of products — i.e., where search costs are low — algorithms tend to coordinate at higher prices. Conversely, when consumers evaluate only a limited number of products, algorithms may drive prices downward by reducing advertising bid costs.

III. Three Risk Scenarios

When examining algorithmic pricing, competition authorities consider three distinct risks.

a. First Scenario: The Hub-and-Spoke Structure

Multiple competitors use the same third-party pricing software. The software pools competitors' confidential commercial data and generates pricing recommendations on that basis. There is no direct contact between competitors, yet the shared software effectively becomes a mechanism for information exchange.

b. Second Scenario: Tacit Collusion

Competitors use different algorithms, but each learns to observe the other's price movements and react accordingly. There is no agreement, no communication — yet the outcome is identical to a cartel output. Traditional competition law theory characterises this as "parallel conduct."

c. Third Scenario: Platform-Driven Coordination

A platform offers its sellers an automated pricing tool. This tool steers sellers toward a price reference set by the platform. When thousands of sellers follow that reference, prices across the platform converge to near-identical levels.

IV. The Algorithmic Pricing Case

In October 2025, the U.S. case In re: RealPage, Inc. Rental Software Antitrust Litigation concluded with a preliminary settlement. Hundreds of landlords and property managers had delegated their rent-setting decisions to RealPage's "RENTmaximizer" software. The software covertly aggregated competitors' rental data to generate "optimal" pricing recommendations. Tenants brought suit, alleging that rents had been artificially inflated as a result.

The U.S. Department of Justice filed suit under Sections 1 and 2 of the Sherman Act. The preliminary settlement established that the parties would not use software that incorporates competitors' confidential data.

A noteworthy aspect of the case is the divergent judicial interpretation of the concept of "agreement":

• In Duffy v. Yardi Systems, multiple competitors' use of the same software was treated as unconditional price coordination.
• In Gibson v. Cendyn Group, competitors' use of the same software under separate contracts was held insufficient, by itself, to constitute evidence of coordination.

Courts have yet to reach consensus, but in March 2025 the DOJ clearly announced that algorithmic price-fixing is actionable under the Sherman Act.

V. The Position in the EU and United Kingdom

The European Commission's stance on "algorithmic pricing" remained largely academic for years. In July 2025, Deputy Director-General for Competition Linsey McCallum confirmed that the Commission is conducting multiple investigations covering algorithmic pricing mechanisms, although the sectors involved have not been publicly disclosed.

From a legal framework perspective, the Commission's 2023 Horizontal Cooperation Guidelines are clear: conduct that is unlawful in the physical world is equally unlawful in the digital world. The use of an algorithm does not shield a company from liability under Article 101 TFEU (anti-competitive agreements) or Article 102 TFEU (abuse of a dominant position).

The UK's Competition and Markets Authority (CMA) has described algorithmic pricing as a "focus area and source of concern." The CMA's Chief Executive has stated that the authority is closely monitoring U.S. proceedings and is examining the link between generative AI risks and algorithmic pricing risks.

VI. A Practical Compliance Framework

Authorities examine three critical risk factors in algorithmic pricing:

• Confidential competitor data: If the software collects competitors' confidential price, rental, or capacity data and generates recommendations on that basis, the risk level is high.
• Ranking tied to reference price: Linking an automated price to seller visibility or ranking within a platform restricts pricing freedom and can give rise to coordination risk.
• The "equalise" function: Fixing prices to a single reference value is risky. Allowing prices to remain slightly above or below that reference is a means of preserving competition.

It is important to understand the data sources used by the software in question, to conduct regular audits, and to obtain contractual guarantees from providers against the use of confidential data. The defence that "we did not write the algorithm, we merely used it" is no longer viable.

*This newsletter is prepared solely for informational purposes and does not constitute any legal advice for a specific case or client.