1. Introduction

Architecture-enabled collusion refers to situations in which the design of a market, platform, software system, communication architecture, information system, contracting structure, or algorithmic environment makes it easier for competing firms to coordinate their conduct and reduce competitive uncertainty.

Traditional cartel law generally focuses on agreements, concerted practices, or coordinated conduct between competitors. Modern markets, however, may create coordination without a conventional meeting or written cartel agreement. A carefully designed architecture can facilitate coordination by:

• making competitors' prices highly visible;
• standardising pricing or bidding mechanisms;
• transmitting competitors' strategic information;
• creating common algorithms or pricing tools;
• automatically responding to competitors' conduct;
• facilitating monitoring of deviations from coordinated behaviour;
• imposing contractual parity or most-favoured-customer clauses;
• reducing the number of meaningful competitive variables;
• creating repeated interactions between competitors; or
• allowing a common intermediary to influence the conduct of several rivals.

The legal question is therefore not simply whether competitors communicated, but also whether the architecture created or facilitated a mechanism through which competitors could align, monitor, or stabilize their market conduct.

2. Meaning of Architecture-Enabled Collusion

The term can be understood through three layers.

A. Physical or institutional architecture

This includes:

• concentrated market structures;
• common ownership arrangements;
• trade associations;
• shared infrastructure;
• common procurement systems;
• centralized bidding platforms;
• joint distribution networks.

B. Digital architecture

This includes:

• pricing algorithms;
• online marketplaces;
• automated auction systems;
• APIs;
• shared databases;
• common software;
• platform ranking systems;
• automated repricing tools.

C. Contractual architecture

This includes:

• resale-price restrictions;
• exclusivity provisions;
• information-sharing clauses;
• MFN/parity clauses;
• common pricing formulas;
• standard contractual terms;
• hub-and-spoke arrangements.

The important point is that architecture is not necessarily unlawful merely because it makes coordination possible. Competition law normally requires the relevant statutory elements—such as an agreement, concerted practice, abuse of dominance, or other prohibited conduct—to be established.

3. Why Architecture Can Facilitate Collusion

Competition between firms normally involves uncertainty:

“What will my competitor do?”

Architecture can reduce this uncertainty.

For example, suppose five competing suppliers independently use a system that displays their current prices almost instantaneously. Each supplier can observe the others and modify its price accordingly.

The system can therefore create:

Transparency → monitoring → rapid retaliation → reduced incentive to compete → stable coordination.

The same mechanism can operate through bidding systems:

Common information → predictable bids → monitoring → punishment of deviations → coordinated tender outcomes.

Thus, architecture can transform an otherwise difficult cartel into a system where coordination becomes easier to sustain.

4. Architecture and the Traditional Concept of Agreement

A major legal distinction is between:

Independent parallel conduct

Competitors independently observe market conditions and adopt similar prices.

This is generally not sufficient, by itself, to establish a cartel.

Coordinated conduct

Competitors knowingly replace independent decision-making with some form of coordination.

This can fall within the concept of an agreement or concerted practice depending upon the jurisdiction.

Architecture-mediated coordination

A third party, platform, algorithm, software provider, association, or common infrastructure may facilitate coordination.

The architecture therefore becomes evidence of the mechanism through which independent competitive decisions may have been replaced or influenced.

5. Hub-and-Spoke Architecture

One of the most important forms is the hub-and-spoke structure.

The structure can be represented as:

Competitor A
↘
Platform / intermediary / hub
↗
Competitor B

The hub may obtain commercially sensitive information from several competitors and transmit information, recommendations, pricing parameters, or strategic signals between them.

The legal difficulty is determining whether:

1. competitors merely used a common intermediary independently;
2. the intermediary facilitated communication between them; or
3. the entire arrangement constituted a coordinated restriction of competition.

This distinction is especially important in digital-platform markets.

6. Algorithmic Architecture

Algorithms can facilitate collusion in several ways.

6.1 Explicit algorithmic coordination

Competitors deliberately program their systems to coordinate prices.

This presents the clearest legal concern.

6.2 Common algorithm

Several competitors use the same pricing algorithm supplied by a third party.

The supplier's role becomes important if the algorithm is designed or used to facilitate coordinated pricing.

6.3 Autonomous learning

Algorithms independently observe competitors and repeatedly react to their conduct.

This creates a more difficult question because the resulting parallel conduct may not necessarily involve a traditional agreement.

6.4 Predictable algorithmic responses

Even without explicit communication, competitors may know that a rival's algorithm will respond automatically to particular pricing changes.

This can make aggressive competition less attractive.

7. Important Case Laws

1. United States v. Apple Inc., 791 F.3d 290 (2d Cir. 2015)

The Apple e-books litigation is a major authority concerning coordinated conduct facilitated through an intermediary.

Apple entered the e-book market and negotiated agency agreements with major publishers. The arrangements included pricing provisions that substantially changed the competitive environment.

The court concluded that Apple had participated in a conspiracy involving the publishers.

Competition-law significance

The case demonstrates that a firm does not necessarily escape cartel liability merely because it does not directly negotiate every aspect of the coordination with every competitor.

A contractual architecture can serve as the mechanism through which coordinated conduct is implemented.

Principle

Contractual and intermediary structures can constitute important evidence of coordination when they facilitate a collective restriction of competition.

8. Eturas UAB v. Lietuvos Respublikos konkurencijos taryba

CJEU, Case C-74/14

This is particularly relevant to digital architecture.

Eturas operated an online travel-booking system used by travel agencies. A system-wide message was transmitted through the platform concerning restrictions on discounts that could be offered through the system.

The Court considered whether travel agencies could be regarded as participating in a concerted practice merely because they received information through the common platform.

The Court held that receipt of a message through the system could be relevant evidence, but participation could not simply be presumed without considering the circumstances and the agencies' conduct.

Significance

The case illustrates how platform architecture can become a channel through which competitively sensitive coordination is communicated.

It is particularly important for:

• online marketplaces;
• common booking systems;
• platform restrictions;
• algorithmic communications; and
• digital information exchanges.

9. AC-Treuhand AG v European Commission

CJEU, Joined Cases C-194/14 P and C-307/18 P

AC-Treuhand concerned the liability of a consultancy acting as a facilitator of cartel arrangements.

The Court accepted that an undertaking which does not itself operate at the same level of the market as the cartel participants may nevertheless incur liability where it intentionally contributes to the implementation of an anticompetitive arrangement.

Significance

This is highly relevant to architecture-enabled collusion because modern coordination may involve:

• software suppliers;
• consultants;
• industry associations;
• data providers;
• platform operators;
• algorithm developers.

The case demonstrates that competition law can reach facilitators, not merely the firms directly fixing prices.

10. In re: RealPage, Inc., Rental Software Antitrust Litigation

The RealPage litigation concerns allegations involving the use of rental-pricing software by landlords and the alleged exchange and use of competitively sensitive information.

The allegations raise important questions about whether a common pricing architecture can facilitate coordinated outcomes among competitors.

Significance

The controversy illustrates a modern competition-law problem:

Can competitors effectively coordinate through a common algorithm even when the firms do not directly communicate with each other?

The legal analysis depends upon evidence concerning information supplied to the system, the operation of the algorithm, the communications surrounding its adoption, and the degree of independent decision-making retained by the users.

11. United States v. Topkins

The Topkins prosecution concerned an alleged agreement among online sellers to coordinate prices for posters and other products using online pricing mechanisms.

The defendants allegedly used algorithms and software to implement pricing arrangements.

Significance

The case is important because it illustrates that:

digital implementation does not change the underlying character of a cartel.

If competitors agree to fix prices, the use of software to implement that agreement does not transform the conduct into lawful independent algorithmic pricing.

The central legal issue remains the existence and nature of coordination.

12. United States v. Socony-Vacuum Oil Co., 310 U.S. 150 (1940)

Although predating digital technology by decades, Socony-Vacuum remains fundamental to understanding architecture-enabled collusion.

The Supreme Court treated agreements among competitors concerning prices as falling within the core prohibition against price fixing.

Significance

The case provides the traditional legal foundation against which modern technological arrangements must be evaluated.

An algorithm, platform, database, or automated system cannot be used to disguise a conventional agreement to fix prices.

13. Interstate Circuit, Inc. v. United States, 306 U.S. 208 (1939)

This is a classic authority on coordinated conduct involving multiple competitors.

The case concerned communications sent to several distributors/exhibitors concerning commercial conditions. The Supreme Court found circumstances supporting an inference of concerted action.

Significance

The case is relevant to architecture-enabled collusion because it illustrates the importance of:

• communications directed simultaneously toward competitors;
• knowledge of the commercial interests of other participants;
• coordinated acceptance of restrictive arrangements; and
• circumstances permitting an inference of concerted conduct.

Modern digital platforms can create similar informational structures much more rapidly.

14. United States v. Container Corporation of America, 393 U.S. 333 (1969)

The case involved exchanges of competitively sensitive information among competitors.

The Supreme Court considered the competitive implications of information exchanges concerning prices.

Significance

This case demonstrates that information architecture itself can be competition-sensitive.

A system that continuously communicates:

• current prices;
• future prices;
• capacity;
• inventories;
• discounts;
• customer allocations;

can potentially reduce strategic uncertainty between competitors.

The competitive assessment therefore cannot focus exclusively on the final price outcome.

15. Indian Competition-Law Perspective

For India, architecture-enabled collusion should principally be analysed under the Competition Act, 2002, particularly the prohibition concerning agreements that cause or are likely to cause an appreciable adverse effect on competition.

Section 3 is particularly relevant to:

• price fixing;
• limiting or controlling production or supply;
• market allocation;
• bid rigging;
• information exchanges;
• arrangements facilitating coordination.

Section 4 becomes relevant where architecture is controlled by a dominant enterprise and is used to exclude competitors or exploit dependent users.

16. Excel Crop Care Ltd. v Competition Commission of India

The Supreme Court's decision in Excel Crop Care is an important Indian authority concerning cartel and bid-rigging principles.

The case concerned alleged coordination in the market for aluminium phosphide tablets supplied to the government.

The Supreme Court considered the nature of cartel conduct and the methodology for imposing penalties.

Relevance

Architecture-enabled collusion in procurement can arise through:

• centralized tender systems;
• predictable tender cycles;
• standardized technical specifications;
• repeated interactions;
• information concerning previous bids;
• bid-submission platforms.

A digital tendering architecture may therefore create both efficiency and opportunities for coordination.

17. Competition Commission of India v. Steel Authority of India Ltd.

The Supreme Court's decision concerning the Competition Commission's investigative and procedural powers is important for understanding the institutional framework within which alleged anticompetitive conduct is investigated.

Relevance

Architecture-enabled collusion may require examination of:

• communications;
• digital records;
• software configurations;
• transaction histories;
• pricing databases;
• internal documents;
• algorithmic instructions.

Competition investigations therefore increasingly intersect with digital evidence.

18. Samir Agarwal v. Competition Commission of India

The Supreme Court's decision concerned allegations relating to online platforms and the interpretation of competition-law concepts in the digital economy.

Although the case was not a pure algorithmic-collusion case, it is relevant to the broader question of how competition law applies to platform-mediated markets.

Significance

Digital architecture can simultaneously create:

• efficiencies;
• lower transaction costs;
• greater transparency;
• network effects;
• information asymmetries; and
• opportunities for exclusion or coordination.

The legal analysis therefore requires examination of the actual market structure and conduct rather than treating technological architecture as inherently pro- or anticompetitive.

19. Architectural Features That Increase Collusion Risk

Certain characteristics deserve particular scrutiny.

None of these features is automatically unlawful. Their significance depends upon the surrounding facts and applicable competition law.

20. Architecture-Enabled Tacit Coordination

A particularly difficult issue is tacit coordination.

Suppose:

1. Firm A raises its price.
2. Firm B's algorithm immediately observes the change.
3. Firm B raises its price.
4. Firm A observes B's response.
5. Both firms maintain the higher price.

There may be no explicit communication.

This creates a fundamental distinction between:

Tacit coordination and concerted practice.

Competition authorities generally cannot simply equate parallel conduct with an unlawful agreement. Evidence must establish the legal elements required by the relevant jurisdiction.

21. The Role of Transparency

Transparency has two opposite effects.

Pro-competitive transparency

It can allow consumers to:

• compare prices;
• identify better offers;
• reduce search costs;
• switch suppliers more easily.

Collusion-facilitating transparency

If competitors receive instantaneous information concerning one another's:

• prices;
• discounts;
• quantities;
• inventories;
• customers;
• future strategies,

transparency can make coordinated conduct easier to sustain.

Therefore:

Consumer transparency and competitor transparency are not necessarily equivalent from a competition-law perspective.

22. Monitoring and Punishment Mechanisms

Successful coordination often requires monitoring.

Architecture can automate monitoring by detecting:

• deviations from agreed prices;
• unusually low bids;
• customer switching;
• deviations from territorial allocation;
• discounts;
• production levels.

An automated system can then trigger:

• price responses;
• withdrawal of discounts;
• exclusion;
• contractual penalties;
• reduced access;
• retaliatory pricing.

This makes the architecture potentially relevant not only to the formation of coordination but also to its stability and enforcement.

23. Hub-and-Spoke Digital Platforms

A platform can occupy a particularly sensitive position.

For example:

Manufacturer A → Platform ← Manufacturer B

If the platform receives confidential pricing information from both manufacturers, the legal analysis may need to consider whether the platform merely provides an independent service or whether it is facilitating coordination.

Relevant evidence may include:

• what information the platform receives;
• whether information is aggregated or individualized;
• whether competitors know the information is shared;
• whether the platform recommends common prices;
• whether competitors communicate through the platform;
• whether the platform monitors compliance;
• whether competitors consciously accepted the coordination mechanism.

24. Common Algorithm Risk

A common algorithm may create several levels of risk.

Level 1 — Neutral tool

Competitors independently use software merely to calculate costs.

Competition concerns are comparatively limited.

Level 2 — Competitively sensitive information

The software collects competitor-specific data.

Greater scrutiny may be required.

Level 3 — Common pricing recommendations

The system recommends prices to multiple competitors using shared information.

The competitive implications become more significant.

Level 4 — Coordinated pricing mechanism

The software is intentionally designed to produce coordinated prices.

This can raise serious cartel concerns.

25. Architecture and Bid Rigging

Procurement markets are particularly vulnerable.

A tender architecture may provide competitors with information concerning:

• previous winning bids;
• bid ranges;
• competitors' participation;
• technical qualifications;
• submission times;
• contract awards.

Repeated tenders can allow firms to develop a predictable pattern such as:

Firm A wins Tender 1 → Firm B wins Tender 2 → Firm C wins Tender 3.

If supported by evidence of coordination, such structures may become relevant to allegations of bid rigging or market allocation.

26. Architecture and Information Exchange

Competition authorities may examine the quality, frequency, timing and granularity of exchanged information.

For example:

Lower strategic sensitivity

Historical, aggregated market statistics.

Higher strategic sensitivity

Current individual competitor prices.

Particularly sensitive

Future prices or commercially confidential strategic plans.

The more precise, current and competitor-specific the information, the greater its potential relevance to coordinated conduct.

27. Competition-Law Compliance for Platform Architects

Businesses designing competitive marketplaces should consider:

1. Data segregation

Competitors should not automatically obtain each other's confidential information.

2. Aggregation

Where commercially appropriate, information can be aggregated so individual firms cannot easily identify competitors' strategies.

3. Access controls

Competitor-specific data should be subject to appropriate restrictions.

4. Algorithm governance

Businesses should document:

• algorithm objectives;
• data sources;
• pricing logic;
• human intervention;
• monitoring procedures.

5. Independent decision-making

Competitors should retain genuine freedom to determine their own commercial strategies.

6. Compliance monitoring

Platforms should assess whether their systems inadvertently facilitate coordinated conduct.

28. Evidence Relevant to Architecture-Enabled Collusion

A competition authority investigating such conduct may examine:

• source code;
• algorithm specifications;
• API logs;
• database architecture;
• access records;
• communications;
• contracts;
• pricing histories;
• bid histories;
• system-generated recommendations;
• internal compliance documents;
• meeting records;
• developer instructions;
• customer data;
• competitor-information flows.

This makes digital forensics increasingly important to competition enforcement.

29. Distinguishing Efficiency from Collusion

Not every common architecture is anticompetitive.

A shared system can generate legitimate efficiencies through:

• reduced transaction costs;
• improved logistics;
• better inventory management;
• fraud prevention;
• interoperability;
• lower search costs;
• improved consumer choice.

The crucial question is therefore not:

“Does the architecture affect competitors?”

but rather:

“How does the architecture affect independent competitive decision-making, and does the resulting conduct satisfy the legal test for an infringement?”

30. Key Legal Tests

A competition-law analysis should generally examine:

A. Relevant market

What products/services and geographic area are affected?

B. Market structure

Consider:

• concentration;
• entry barriers;
• switching costs;
• network effects;
• frequency of transactions.

C. Architecture

Identify the actual technical, contractual or institutional structure.

D. Information flow

Determine:

• who receives information;
• what information;
• when;
• at what level of detail.

E. Decision-making

Ask whether firms retain genuine independent pricing and commercial decisions.

F. Intent and knowledge

Examine evidence concerning the parties' knowledge and purpose where legally relevant.

G. Effects

Consider:

• prices;
• output;
• innovation;
• quality;
• consumer choice;
• entry;
• foreclosure.

H. Legal classification

Determine whether the conduct constitutes:

• an agreement;
• concerted practice;
• cartel;
• information exchange;
• hub-and-spoke coordination;
• abuse of dominance;
• vertical restraint; or
• another competition-law infringement.

31. Major Case-Law Principles at a Glance

32. Emerging Risks from AI Agents

The problem becomes more complex with autonomous AI agents.

Imagine competing companies deploying AI agents that independently:

• monitor market prices;
• predict competitors' responses;
• change prices;
• negotiate contracts;
• adjust output;
• react to demand;
• monitor deviations.

If each agent learns that aggressive price competition triggers immediate retaliation, the market could potentially develop stable coordination without conventional human meetings.

This raises difficult questions about:

1. attribution of algorithmic decisions;
2. human supervision;
3. foreseeability;
4. evidence of communication;
5. common algorithmic architecture;
6. intentional facilitation;
7. autonomous tacit coordination; and
8. appropriate remedies.

The technology does not eliminate established competition-law principles, but it can make proving how coordination occurred considerably more difficult.