Competition Law and Algorithmic Coordination Infrastructures

Competition Law and Algorithmic Coordination Infrastructures

1. Introduction

Algorithmic coordination infrastructures are digital systems that allow firms to collect, process, exchange, or react to market information through software. They can include pricing algorithms, common revenue-management systems, online marketplaces, data-sharing platforms, automated booking systems, benchmarking services, and AI-based decision tools.

From a competition-law perspective, the technology itself is normally not unlawful. The concern arises when the infrastructure makes it easier for competitors to coordinate prices, output, discounts, capacity, customers, or other commercially important decisions.

Traditional cartels may depend on meetings, telephone calls, or direct exchanges between competitors. Digital markets can operate differently. A common algorithm or intermediary may collect confidential information from many competing firms and then provide recommendations or signals based on that information. This can potentially reduce the uncertainty that normally exists between independent competitors.

Competition authorities therefore examine the economic function of the infrastructure rather than simply asking whether human competitors directly communicated.

2. Meaning of Algorithmic Coordination Infrastructure

An algorithmic coordination infrastructure can be understood as a technological arrangement that connects the commercial decision-making processes of multiple market participants.

A simplified structure is:

Competitor A → Data → Common Platform → Algorithm → Recommendation

Competitor B → Data → Common Platform → Algorithm → Recommendation

Competitor C → Data → Common Platform → Algorithm → Recommendation

The platform may therefore occupy a central position between otherwise competing businesses.

For example, competing hotels might independently subscribe to the same revenue-management system. Each hotel supplies information concerning occupancy, prices or demand. The software processes the information and recommends room prices.

This does not automatically establish a cartel. However, competition concerns become stronger where the system uses current or future confidential information from competitors and participants understand that their rivals are using the same mechanism.

3. Main Competition-Law Issues

A. Algorithmic Price Coordination

The clearest concern is coordinated pricing.

Suppose several competing businesses use the same pricing system. The system receives non-public information from those businesses and recommends similar prices.

If the arrangement effectively replaces independent price determination with coordinated decision-making, competition authorities may investigate whether an unlawful agreement or concerted practice exists.

The important point is that competition law generally focuses on the substance of coordination. Using software as an intermediary does not automatically remove competition-law responsibility.

B. Exchange of Competitively Sensitive Information

Algorithms depend heavily on data.

Information concerning future prices, production, capacity, discounts, costs, customers or strategic plans can be particularly sensitive.

When competitors receive detailed information about one another, uncertainty in the market can decline. Firms may become better able to predict how competitors will behave and adapt their own conduct accordingly.

Historical and sufficiently aggregated information normally presents different risks from detailed, recent or forward-looking company-level information.

C. Hub-and-Spoke Coordination

Algorithmic systems can also create a modern form of hub-and-spoke arrangement.

The software provider acts as the hub, while competing users are the spokes.

The legal question becomes whether the competitors have merely entered separate agreements with a technology supplier or whether the surrounding facts establish broader coordination between the competitors through that supplier.

D. Monitoring and Enforcement

Successful coordination may require participants to detect deviations.

Algorithms can make this easier because they can continuously observe prices and other market variables. If one participant reduces its price, competing systems may react rapidly.

Consequently, technology can potentially strengthen an existing coordinated arrangement by improving monitoring and reducing the benefit of secretly departing from it.

E. Market Transparency

Transparency has mixed competitive effects.

Public price transparency can benefit consumers by making comparison easier. However, private exchanges of detailed strategic information among competitors can produce different effects.

Competition authorities therefore distinguish between transparency available to consumers generally and confidential transparency created primarily among competing suppliers.

4. Important Case Laws and Enforcement Proceedings

Case 1: United States v. David Topkins

Jurisdiction: United States
Court: U.S. District Court for the Northern District of California
Year: 2015

This is one of the earliest major U.S. criminal matters involving algorithm-assisted price fixing.

The case concerned posters sold through an online marketplace. The U.S. Department of Justice alleged that competing sellers agreed to fix prices and used pricing algorithms to implement their agreement.

The algorithm was therefore not treated as an independent actor responsible for the conduct. Rather, it was a technological mechanism through which an alleged human agreement was implemented.

Topkins pleaded guilty to the charged price-fixing conspiracy.

Importance

The case establishes an important principle for algorithmic competition law:

An unlawful price-fixing agreement does not become lawful merely because software implements it.

Traditional cartel principles can therefore apply when competitors intentionally program or employ algorithms to carry out their agreement.

Case 2: Eturas UAB and Others v Lithuanian Competition Council — C-74/14

Court: Court of Justice of the European Union
Judgment: 21 January 2016

Eturas concerned Lithuanian travel agencies participating in a common online booking system.

A system administrator sent an electronic message concerning restrictions on discounts available through the booking platform, and the system technically implemented a restriction on discount rates.

The Court considered when travel agencies using a shared computerized system could be regarded as participating in a concerted practice.

Knowledge was important. Participation could not simply be established because a company happened to use the platform. Evidence concerning awareness of the communication and the participant's conduct remained significant, while businesses had to have an effective opportunity to rebut relevant presumptions.

Importance

Eturas is particularly significant for algorithmic infrastructures because it demonstrates that a common digital platform can become the mechanism through which coordination occurs.

At the same time, use of common technology alone does not automatically establish liability.

Case 3: In re RealPage, Inc., Rental Software Antitrust Litigation

Jurisdiction: United States
Court: U.S. District Court, Middle District of Tennessee

Private plaintiffs brought litigation concerning revenue-management software used by landlords.

The allegations focus on whether competing property managers supplied commercially sensitive rental information to a common software provider and received algorithmically generated pricing recommendations.

The U.S. government has also filed statements addressing legal principles relevant to algorithmic price-fixing allegations. Government filings emphasize that algorithmic tools do not require a completely new body of antitrust law simply because the alleged coordination is technologically sophisticated.

Importance

RealPage litigation illustrates the modern common-algorithm problem:

Can separate competitors' participation in a shared pricing infrastructure amount to concerted action when the infrastructure pools sensitive information and generates pricing recommendations?

The answer depends on evidence concerning agreement, knowledge, information exchange and the actual operation of the system rather than merely the existence of software.

Case 4: United States and Plaintiff States v. RealPage, Inc.

Jurisdiction: United States
Filed: 2024

This government enforcement proceeding is distinct from the private RealPage litigation.

The U.S. Department of Justice and participating states alleged violations of Sections 1 and 2 of the Sherman Act involving RealPage's revenue-management technology. According to the government's complaint, competing landlords supplied non-public information concerning rental rates and other leasing terms to RealPage's algorithmic pricing system.

The allegations should be distinguished from established judicial findings: filing an antitrust complaint does not itself prove the allegations.

The proceeding is nevertheless highly important because it directly addresses a centralized algorithmic infrastructure serving numerous competitors.

Importance

It demonstrates how authorities may examine:

• collection of competitors' confidential data;
• centralized algorithmic recommendations;
• competitor knowledge of the common system;
• acceptance or implementation of recommendations; and
• the software provider's role in the competitive process.

It is therefore one of the clearest contemporary examples of enforcement focused specifically on algorithmic coordination infrastructure.

Case 5: Cornish-Adebiyi v. Caesars Entertainment, Inc.

Jurisdiction: United States
Industry: Hotels

This litigation concerned allegations involving hotel operators and common revenue-management technology.

In 2024, the U.S. Federal Trade Commission and Department of Justice filed a statement of interest explaining their interpretation of Section 1 principles applicable to algorithmic price-fixing allegations.

The agencies argued that direct communications between competitors are not necessarily required to plead an agreement where an intermediary allegedly facilitates coordination. They also emphasized that conduct that would violate competition law when performed through human interaction does not become permissible simply because an algorithm performs the relevant function.

Importance

The case illustrates the potential hub-and-spoke architecture of algorithmic coordination.

A software company may operate at the center while numerous competing businesses use the same infrastructure.

The important legal question remains whether sufficient facts establish an agreement or concerted action rather than merely parallel use of commercially available technology.

Case 6: Gibson v. Cendyn Group, LLC

Jurisdiction: United States
Industry: Hotel revenue-management software

This litigation also concerns the relationship between common pricing technology and Section 1 of the Sherman Act.

The U.S. Department of Justice submitted an amicus brief explaining its position on how existing antitrust principles apply to pricing algorithms. The government described algorithms as another technological mechanism through which competitors may potentially restrict competition, rather than something automatically outside traditional antitrust doctrine.

Importance

The litigation raises an important distinction between:

independent adoption of similar software

and

concerted participation in a common mechanism for coordinating competitive decisions.

The existence of a common vendor by itself is therefore different from evidence establishing an agreement among the competing users.

Case 7: United States v. Agri Stats, Inc.

Jurisdiction: United States
Filed: 2023

Agri Stats is not primarily an AI pricing-algorithm case, but it is extremely relevant to algorithmic coordination infrastructures because it concerns a centralized data and benchmarking intermediary.

The Department of Justice alleged that Agri Stats collected detailed information concerning prices, costs and output from competing meat processors and distributed reports that allowed participants to compare their operations with rivals.

In May 2026, the DOJ announced a proposed settlement restricting important aspects of these information exchanges, including non-public pricing information and certain company- or facility-level production, cost and labor data.

Importance

Agri Stats demonstrates that competition concerns do not depend on sophisticated AI.

A centralized information infrastructure itself can potentially facilitate coordination by reducing strategic uncertainty between competitors.

This principle is directly relevant when modern algorithms automate the collection, analysis and distribution of the same kinds of commercially sensitive information.

5. Legal Tests Applied to Algorithmic Coordination

Courts and competition authorities generally do not treat "algorithmic collusion" as a completely separate offence. Existing rules governing agreements, concerted practices and information exchanges remain central.

Several questions become especially important.

Agreement or Concerted Practice

Authorities must normally identify sufficient evidence connecting the competitors' behavior.

Mere parallel pricing is generally different from an agreement.

Where an algorithmic intermediary exists, investigators may therefore examine communications, contracts, system design, data-sharing arrangements and evidence concerning what participating firms understood about the system.

Knowledge

Knowledge can be particularly important.

A company unknowingly using software that independently produces a market price presents a different legal situation from competitors knowingly participating in an infrastructure designed to pool sensitive information and coordinate commercial decisions.

Eturas illustrates the importance of this distinction.

Nature of Data

Authorities may examine whether information is:

• public or confidential;
• historical or forward-looking;
• aggregated or company-specific;
• old or near real-time;
• general or commercially strategic.

Current company-level pricing or output information usually raises more significant competition concerns than old, aggregated statistics.

Market Characteristics

Algorithmic coordination may be easier in markets containing relatively few competitors, standardized products, frequent transactions and highly observable prices.

Algorithms can potentially increase market transparency and speed up competitive reactions.

However, similar prices do not by themselves prove unlawful coordination. Comparable algorithms can independently respond to identical market conditions and reach similar results.

6. Autonomous Algorithmic Coordination

One of the hardest theoretical problems arises where algorithms learn to behave in parallel without an express agreement between their operators.

For example, several independently designed AI systems might discover that aggressive price competition lowers profits and consequently adopt strategies producing relatively stable prices.

Traditional competition law faces difficulties here because cartel provisions normally require some form of agreement, coordination or concerted practice.

Purely unilateral but parallel algorithmic behavior may therefore fall outside traditional cartel prohibitions unless additional evidence establishes the legally required connection between the businesses.

This is why authorities focus heavily on human decisions surrounding the algorithm—who designed it, what data it receives, what competitors know, whether a common provider is involved, and whether businesses agreed to participate in the mechanism.

7. Compliance Implications

Businesses using common pricing, benchmarking or AI infrastructure should preserve independent commercial decision-making.

Important competition-law safeguards include limiting unnecessary exchanges of competitors' confidential information, carefully controlling access to current or forward-looking strategic data, independently reviewing algorithmic recommendations, maintaining auditable records concerning the operation of pricing systems, and conducting competition-law assessment before implementing shared industry platforms.

Companies should also remember that outsourcing a commercial decision to software does not necessarily outsource legal responsibility.

8. Conclusion

Algorithmic coordination infrastructures represent the intersection of traditional competition law and modern digital markets.

The central problem is not simply whether firms use algorithms. The more important questions are how competitors are connected through the infrastructure, what information enters the system, what information or recommendations emerge from it, what participating firms know, and whether their commercial decisions remain genuinely independent.

Cases and proceedings such as United States v. Topkins, Eturas, In re RealPage, United States v. RealPage, Cornish-Adebiyi v. Caesars Entertainment, Gibson v. Cendyn Group, and United States v. Agri Stats demonstrate different parts of this developing framework.

Together, they show that existing concepts—including agreements, concerted practices, hub-and-spoke arrangements and exchanges of competitively sensitive information—can apply to digital coordination. At the same time, competition law must distinguish unlawful coordination from lawful independent use of similar technology.

Therefore, the fundamental competition-law principle remains largely unchanged: each competitor should determine its competitive strategy independently rather than replacing competitive uncertainty with a shared mechanism for coordination.