1. Introduction

Autonomous coordination networks are technological or organizational systems in which software agents, algorithms, artificial-intelligence systems, smart contracts, or interconnected digital platforms independently observe market conditions and adjust conduct without continuous human intervention.

Examples include:

• AI-powered pricing systems;
• algorithmic procurement systems;
• autonomous trading platforms;
• smart-contract networks;
• decentralized marketplaces;
• advertising-auction systems;
• logistics optimization networks;
• energy-management platforms;
• autonomous purchasing systems; and
• competing firms using interconnected pricing algorithms.

The competition-law challenge arises because autonomous systems can potentially produce coordinated market outcomes without traditional human communications.

The fundamental question is:

When does autonomous algorithmic coordination become legally attributable to the firms using or controlling the systems?

The answer differs across jurisdictions, but traditional concepts such as agreement, concerted practice, cartel, information exchange, hub-and-spoke coordination, conscious parallelism, and unilateral algorithmic conduct remain important.

2. Meaning of Autonomous Coordination Networks

An autonomous coordination network can be understood as:

A network of software agents or economic actors capable of independently observing market information, processing that information, and modifying commercial behaviour according to programmed or machine-learning objectives.

A simplified structure is:

Firm A → Algorithm A
Firm B → Algorithm B
Firm C → Algorithm C

The algorithms continuously observe:

• competitors' prices;
• inventory;
• demand;
• bids;
• capacity;
• advertising rates;
• consumer behaviour;
• market conditions.

They then autonomously change commercial decisions.

For example:

Competitor price increases → Algorithm detects increase → Algorithm raises own price → Competitor's algorithm observes change → Competitor raises price → repeated interaction → stable high-price outcome.

The legal difficulty is determining whether this is:

1. lawful independent adaptation;
2. tacit coordination;
3. a concerted practice;
4. an agreement facilitated by technology; or
5. unilateral conduct by a dominant undertaking.

3. Why Competition Law Is Concerned

Competition law traditionally assumes that collusion involves human decision-makers.

Traditional cartel:

Company A executives ↔ Company B executives → agreement → higher prices

Autonomous coordination may look different:

Algorithm A ↔ market data ↔ Algorithm B → coordinated outcome

There may be no email stating:

"We agree to raise prices."

Nevertheless, algorithms may be designed, trained, or configured to respond predictably to competitors.

The main competition concerns are therefore:

• price coordination;
• output restriction;
• market allocation;
• bid coordination;
• exchange of commercially sensitive information;
• coordinated advertising;
• algorithmic signalling;
• hub-and-spoke arrangements;
• exclusionary coordination;
• autonomous collusion; and
• facilitation of traditional cartels.

4. Traditional Cartel Law and Autonomous Systems

The basic cartel categories remain relevant:

Price fixing

Algorithms coordinate prices directly or indirectly.

Output restriction

Algorithms reduce supply or production.

Market allocation

Software divides customers or geographic territories.

Bid rigging

Automated systems coordinate tender bids.

Information exchange

Algorithms continuously exchange commercially sensitive information.

Collective refusal

Autonomous systems identify and exclude particular suppliers or customers.

Technology changes the mechanism, but not necessarily the underlying competitive harm.

5. Agreement Versus Autonomous Behaviour

This is the central doctrinal issue.

Competition law generally requires some form of:

• agreement;
• concerted practice;
• communication;
• coordination; or
• legally relevant unilateral conduct,

depending upon the jurisdiction.

A purely autonomous algorithm that independently reacts to publicly available information may not automatically constitute an unlawful agreement.

For example:

Firm A uses Algorithm A and Firm B uses Algorithm B. Both algorithms independently observe publicly posted prices and adjust prices.

The fact that both prices increase does not automatically prove a cartel.

The authorities would need to examine whether there was:

• communication;
• common algorithmic design;
• coordination instructions;
• exchange of confidential information;
• facilitating conduct;
• common software provider;
• human involvement; or
• other evidence of concerted behaviour.

6. Conscious Parallelism

Conscious parallelism occurs when firms independently adopt similar commercial strategies because they observe each other's conduct.

For example:

• Firm A raises price;
• Firm B observes it;
• Firm B raises price;
• Firm A observes B;
• Firm A maintains its higher price.

This may produce parallel prices without an express agreement.

Generally, parallel behaviour alone is not sufficient to establish a cartel in many legal systems.

The problem becomes more complicated when algorithms make such reactions:

• faster;
• more predictable;
• continuous;
• automated; and
• extremely sensitive to competitor conduct.

7. Autonomous Learning Algorithms

Machine-learning systems create an additional complication.

Traditional software follows predetermined instructions:

"If competitor price falls by 5%, reduce our price by 5%."

AI systems may instead identify strategies through training and repeated market interaction.

They can potentially learn:

• when to increase prices;
• when to reduce discounts;
• when to avoid aggressive competition;
• how competitors respond;
• when to punish deviations.

This raises the possibility of emergent coordination.

However, the mere emergence of parallel conduct should not automatically be equated with an unlawful agreement.

8. Case Law

Case 1: United States v. Airline Tariff Publishing Co.

Facts

Airlines used an electronic tariff publishing system that allowed carriers to communicate fare information.

The U.S. Department of Justice challenged practices involving advance communication of fare changes.

Competition issue

The case demonstrated how information technology can facilitate coordination without requiring traditional face-to-face meetings.

Significance

Electronic information systems can function as mechanisms through which competitors:

• signal future prices;
• monitor competitors;
• discipline deviations; and
• coordinate commercial behaviour.

Relevance to autonomous coordination networks

Modern algorithms can perform these functions much faster and more systematically.

The case therefore provides an important conceptual foundation for analysing technology-enabled coordination.

9. Case 2: United States v. Apple Inc.

Facts

The U.S. government challenged Apple's alleged role in coordinating e-book pricing with publishers.

The case involved communications among publishers and Apple's contractual arrangements.

Competition issue

The case demonstrates that a technological platform can become a hub through which competitors coordinate conduct.

Hub-and-spoke significance

The structure can be represented as:

Publisher A → Apple ← Publisher B

If the central platform facilitates coordination among competing firms, the arrangement may attract antitrust scrutiny.

Relevance

Autonomous coordination networks may create a technologically sophisticated version of the same problem:

Competitor A → common algorithm/platform ← Competitor B

The legal issue remains whether the intermediary knowingly facilitates coordination or merely provides neutral infrastructure.

10. Case 3: United States v. Topkins

Facts

Topkins and other sellers participated in an alleged conspiracy involving online retail pricing.

The participants used pricing algorithms to implement an agreement to maintain prices.

Competition issue

The case is particularly significant because it involved algorithmic pricing in an online marketplace.

The defendants allegedly used software to monitor and implement coordinated pricing.

Significance

The case demonstrates an important principle:

An algorithm does not eliminate antitrust responsibility merely because software performs the final pricing action.

If human actors agree to coordinate and then use software to implement that agreement, the automated character of implementation does not transform the conduct into lawful independent competition.

11. Case 4: Eturas v. Lietuvos Respublikos Konkurencijos Taryba

Facts

Several travel agencies used the same online booking system.

The platform administrator transmitted a message through the system concerning limits on discounts.

The issue was whether the travel agencies had participated in a concerted practice.

European Court of Justice

The Court examined whether knowledge of the electronic message, combined with continued participation in the system, could establish involvement in a concerted practice.

Significance

The case is extremely important for digital competition law.

It demonstrates that:

• electronic communications can constitute evidence of coordination;
• participation through a digital platform can have antitrust consequences;
• firms cannot necessarily avoid responsibility merely because coordination occurs through software.

Autonomous-network relevance

Modern platforms can communicate commercial instructions electronically without conventional meetings.

The legal inquiry therefore increasingly focuses on what the system communicated, what firms knew, and how they responded.

12. Case 5: T-Mobile Netherlands v. Raad van bestuur van de Nederlandse Mededingingsautoriteit

Facts

Several mobile telecommunications operators participated in a meeting where commercially sensitive information was discussed.

Legal issue

The European Court of Justice examined whether the exchange of strategically sensitive information could constitute a concerted practice.

Principle

Competition law can treat the exchange of strategic information as reducing uncertainty about competitors' future conduct.

Relevance to autonomous networks

Algorithms constantly process information.

If autonomous systems receive:

• future pricing information;
• capacity information;
• customer information;
• bidding strategies; or
• commercially sensitive forecasts,

the resulting coordination can potentially be more powerful than conventional information exchanges.

The case therefore illustrates the legal importance of strategic uncertainty.

13. Case 6: AC-Treuhand v European Commission

Facts

AC-Treuhand acted as a service provider supporting cartel participants in the chemical industry.

The European Union courts considered whether a business that was not itself operating in the cartelized product market could nevertheless incur liability for facilitating the cartel.

Significance

The case expanded the conceptual understanding of participation in cartel arrangements.

Autonomous-network relevance

Modern coordination may involve:

• cloud providers;
• algorithm suppliers;
• data intermediaries;
• software providers;
• platform operators.

The AC-Treuhand doctrine is therefore relevant when considering whether an intermediary knowingly facilitates anticompetitive coordination.

The precise liability of a technology provider depends on the applicable jurisdiction and the evidence of its involvement.

14. Case 7: Competition and Markets Authority — Online Hotel Booking

The UK hotel-booking investigations involving online platforms and hotel providers provide an important background to digital coordination.

The investigations examined online pricing restrictions and the use of booking platforms.

The competition concerns included:

• price parity;
• platform restrictions;
• restrictions on discounting;
• information transparency; and
• reduced competition among booking platforms.

Relevance to autonomous systems

Where algorithms monitor hotel prices across platforms, parity arrangements can become automatically enforceable.

An algorithm may detect:

Hotel offers lower price elsewhere → platform detects deviation → platform responds or imposes consequences.

This can make enforcement of potentially restrictive contractual arrangements instantaneous.

15. Case 8: Google Search (Shopping)

The European Commission's Google Shopping decision provides a different but relevant form of algorithmic competition analysis.

Google's search algorithm determined how competing services were displayed to consumers.

The competition concern was not cartel coordination but algorithmic leveraging and self-preferencing.

This illustrates that autonomous algorithms can create competition problems in two different ways:

Coordination

Algorithms may facilitate coordination between competitors.

Exclusion

An algorithm controlled by a dominant platform may disadvantage competitors.

Thus:

Algorithmic competition problems are not limited to cartels.