Competition Law and Intelligent Institutional Coordination Systems

1. Introduction

Intelligent Institutional Coordination Systems (IICS) may be understood as digital or technology-enabled systems through which multiple institutions, enterprises, regulators, associations, platforms, intermediaries, or market participants coordinate information, standards, prices, access conditions, allocation decisions, compliance processes, or strategic responses.

Examples include:

• common digital procurement systems;
• industry-wide data exchanges;
• algorithmic pricing platforms;
• trade-association information systems;
• shared logistics or allocation platforms;
• common AI decision-making systems;
• industry standards and interoperability platforms;
• digital marketplaces connecting competing suppliers;
• common risk-assessment or credit systems;
• shared infrastructure and API ecosystems.

The competition-law problem arises when a system designed to improve institutional coordination also reduces the independence of competing undertakings.

Modern competition law therefore increasingly examines not merely whether competitors communicated directly, but whether a common system allowed them to obtain, process, transmit, or act upon competitively sensitive information in a coordinated manner. The OECD's 2026 work specifically notes that digital tools, common platforms, data intermediaries and algorithms can make information exchanges more frequent, granular and actionable.

2. Meaning of Intelligent Institutional Coordination Systems

An IICS normally contains five components:

A. Institutional participants

Several independent entities participate in a common system.

Examples:

• competing banks;
• hospitals;
• insurers;
• logistics companies;
• landlords;
• manufacturers;
• retailers;
• government agencies;
• trade associations.

B. Common information architecture

The system collects information from participants and may process:

• prices;
• costs;
• capacity;
• inventories;
• customers;
• demand forecasts;
• discounts;
• output;
• strategic plans;
• market shares.

C. Intelligence layer

The system may employ:

• AI;
• machine learning;
• predictive analytics;
• automated pricing;
• optimisation algorithms;
• recommendation engines;
• scoring systems.

D. Coordination mechanism

The system can facilitate:

• common standards;
• allocation;
• price alignment;
• supply coordination;
• market segmentation;
• access decisions;
• customer allocation.

E. Institutional governance

A central administrator may establish:

• participation rules;
• data-access rules;
• technical standards;
• algorithms;
• compliance procedures;
• dispute-resolution mechanisms.

The institutional coordinator itself may therefore become competition-law relevant.

3. Central Competition-Law Principle

The basic principle is:

Competition law protects independent competitive decision-making.

Coordination is not automatically unlawful. Legitimate coordination may produce substantial efficiencies—for example, interoperability, safety, standardisation, infrastructure sharing or improved logistics.

The legal concern arises where coordination:

1. removes uncertainty between competitors;
2. facilitates exchange of competitively sensitive information;
3. coordinates prices or output;
4. enables market allocation;
5. restricts independent commercial decision-making;
6. creates discriminatory access;
7. facilitates exclusion of rivals; or
8. allows an intermediary to become a mechanism for cartel coordination.

The OECD's 2026 analysis emphasises that information sharing can produce efficiencies but can also facilitate tacit or explicit coordination, particularly where digital systems make information more granular and actionable.

4. Relevant Legal Framework

A. India

The principal provisions are:

Section 3(1), Competition Act, 2002

Prohibits agreements, arrangements or understandings that cause or are likely to cause an appreciable adverse effect on competition.

Section 3(3)

Particularly important where enterprises or associations of enterprises coordinate:

• prices;
• production;
• supply;
• markets;
• sales;
• purchases.

Section 3(4)

Relevant to vertical coordination involving:

• exclusive dealing;
• resale-price maintenance;
• refusal to deal;
• tying;
• exclusive distribution.

Section 4

Relevant where an intelligent coordination infrastructure is controlled by a dominant enterprise and is used for:

• discriminatory access;
• self-preferencing;
• exclusion;
• leveraging;
• denial of interoperability;
• excessive or unfair conditions.

5. Intelligent Coordination and Information Exchange

The first major competition concern is information exchange.

Not all information exchange is prohibited. The risk depends upon factors such as:

• whether information is competitively sensitive;
• whether it is current or historical;
• whether it is public or private;
• frequency of exchange;
• level of aggregation;
• market concentration;
• transparency of the market;
• purpose of the exchange;
• whether competitors can identify each other's conduct.

Information concerning future prices, output, costs, capacity and strategic plans is generally more competitively sensitive than historical, aggregated information.

6. Hub-and-Spoke Coordination

An IICS can create a hub-and-spoke structure.

Traditional model

Competitor A → Hub ← Competitor B

The hub receives sensitive information from A and B and facilitates coordination.

Intelligent model

A → AI/Data Platform ← B
↓
Common algorithm
↓
Coordinated commercial recommendations

The important question becomes whether the technology merely provides a neutral service or actually facilitates coordinated conduct.

7. Six Major Case Laws

Case 1: T-Mobile Netherlands BV v Raad van Bestuur van de Nederlandse Mededingingsautoriteit

Court: Court of Justice of the European Union
Case: C-8/08
Year: 2009

Facts

A group of competing mobile telecommunications operators participated in a meeting at which commercially sensitive information concerning matters such as dealer remuneration and market strategy was discussed.

Principle

The CJEU held that certain exchanges of competitively sensitive information can constitute a restriction of competition by object, depending upon their nature and context.

Relevance to IICS

The important lesson is that a sophisticated coordination system does not have to contain an explicit agreement saying:

"We agree to fix prices."

A mechanism that substantially reduces strategic uncertainty between competitors can itself create serious competition-law exposure.

Application

An AI-enabled industry system that continuously communicates:

• future pricing intentions;
• capacity;
• strategic discounts;
• customer targeting;

could potentially create the same underlying competition concern in a technologically different form.

8. Case 2: Eturas UAB and Others v Lietuvos Respublikos konkurencijos taryba

CJEU: Case C-74/14
Judgment: 21 January 2016

This is one of the most directly relevant cases to intelligent digital coordination.

Facts

Travel agencies used a common online booking system operated by Eturas.

The system administrator transmitted a message indicating that online discounts would be restricted to 3%, and the technical system implemented the restriction.

Legal issue

Could participation in the common electronic system contribute to a finding of concerted practice?

Holding

The CJEU recognised that a common digital platform can transmit a coordinating signal. However, participation in the system alone does not automatically establish liability. Evidence concerning awareness and acceptance was important.

The OECD identifies Eturas as a significant example of a digital system transmitting a coordination signal and operationalising a common commercial parameter.

IICS significance

This case establishes an important principle:

Technology does not make an otherwise problematic coordination mechanism immune from competition law.

The relevant questions include:

• Who controlled the system?
• What message was transmitted?
• Did participants know about it?
• Did they accept the system?
• Did they distance themselves?
• Did the system actually affect commercial conduct?

9. Case 3: Dole Food Company, Inc. v European Commission

CJEU: Joined Cases C-286/13 P and related appeals
Year: 2015

Facts

Banana producers/importers exchanged commercially sensitive information concerning pricing.

Principle

The European courts examined whether exchanges of information concerning future pricing intentions could reduce uncertainty and facilitate coordination.

Relevance to intelligent systems

A common AI system could potentially accomplish indirectly what competitors cannot lawfully accomplish through direct communication.

For example:

Competitor A uploads future pricing data → AI system processes it → Competitor B receives commercially useful information.

The absence of an email or meeting does not necessarily eliminate the competition concern.

Legal lesson

Competition analysis focuses on the economic and competitive function of the communication, not merely its technological form.

10. Case 4: AC-Treuhand AG v European Commission

CJEU: Case C-194/14 P
Year: 2015

Facts

AC-Treuhand acted as an intermediary in cartel arrangements involving manufacturers.

Principle

The case significantly broadened the understanding of participation in cartel conduct by recognising that an undertaking can face competition-law consequences even where it is not itself a conventional producer or seller in the affected market.

IICS significance

This is extremely important for:

• AI providers;
• data intermediaries;
• industry platforms;
• software providers;
• trade associations;
• algorithmic coordinators.

An entity cannot necessarily avoid competition-law scrutiny simply by saying:

"I do not sell the product; I only operate the coordination infrastructure."

Where an intermediary knowingly facilitates anticompetitive coordination, its role may become legally significant.

11. Case 5: Samir Agrawal v Competition Commission of India

Supreme Court of India: 2020

This is an important Indian case concerning algorithmic pricing and hub-and-spoke allegations.

Facts

The informant alleged that Ola and Uber's algorithms facilitated price coordination between drivers.

The CCI rejected the allegation at the prima-facie stage, and the matter ultimately reached the Supreme Court.

The Indian decision distinguished conventional hub-and-spoke arrangements from the operation of cab-aggregator algorithms. The reasoning noted that the existence of algorithmically determined prices did not, without more, establish collusion between drivers.

Importance

The case demonstrates an important limitation:

Algorithmic price similarity is not automatically proof of collusion.

Competition authorities must establish the legally relevant elements of coordination.

IICS lesson

A common algorithm can exist for legitimate reasons:

• demand prediction;
• congestion management;
• matching;
• dynamic pricing;
• resource allocation.

The critical issue is whether the system merely optimises independently or instead facilitates concerted conduct among competitors.

12. Case 6: United States v RealPage, Inc.

U.S. Department of Justice antitrust proceeding, initiated 2024

This is a particularly important modern example of algorithmic coordination.

The DOJ alleged that competing landlords provided competitively sensitive information to RealPage, whose pricing software used that information to generate rental recommendations. The DOJ alleged violations of Sections 1 and 2 of the Sherman Act.

The litigation has subsequently involved multiple landlord defendants and proposed settlements concerning algorithmic coordination and information sharing.

Why it matters

The RealPage proceedings demonstrate the modern form of the hub-and-spoke problem:

Landlord A
↓
Competitively sensitive data
↓
Common algorithmic platform
↓
Pricing recommendation
↑
Competitively sensitive data
↑
Landlord B

The alleged coordination does not require competitors to sit in a room and agree upon a price.

Important qualification

RealPage should be distinguished from a final appellate judicial precedent. It is an ongoing antitrust enforcement proceeding and settlement process, rather than a final Supreme Court determination establishing all of the allegations as adjudicated facts.

13. Comparative Case-Law Table

14. Institutional Coordination Versus Cartel Coordination

A crucial distinction must be maintained.

Legitimate institutional coordination

Examples:

• common safety standards;
• interoperable technical standards;
• cybersecurity protocols;
• infrastructure sharing;
• common emergency procedures;
• anonymised statistical databases;
• regulatory compliance systems.

These may generate substantial efficiencies.

Potentially problematic coordination

Examples:

• common future-price database;
• common algorithm using rivals' confidential pricing;
• coordinated output restrictions;
• allocation of customers;
• exclusionary access rules;
• collective refusal to supply;
• common algorithm designed to suppress competitive discounts.

The same technology can therefore be procompetitive in one configuration and anticompetitive in another.

15. Algorithmic Coordination

Intelligent systems introduce several forms of coordination.

A. Explicit algorithmic coordination

Competitors intentionally configure a system to produce a coordinated result.

B. Information-mediated coordination

Competitors provide sensitive data to a common intermediary.

C. Algorithmic monitoring

The system continuously observes competitors and immediately responds to their prices.

D. Predictive coordination

AI predicts competitors' likely actions and automatically adjusts the firm's conduct.

E. Common-parameter coordination

Different firms use identical software containing parameters that suppress competitive differentiation.

F. Autonomous coordination

Algorithms independently react to market signals and produce parallel outcomes without direct human communication.

The last category creates difficult questions concerning intent, foreseeability, knowledge and attribution.

The OECD's current analysis specifically identifies machine-mediated inference, common pricing tools and shared data infrastructure as emerging forms of coordination that may challenge traditional information-exchange concepts.

16. Competition Risks Created by IICS

16.1 Price Coordination

A common intelligent system may recommend similar prices to competing enterprises.

Risk increases where the system uses:

• rivals' current prices;
• future pricing intentions;
• confidential discounts;
• individualised pricing information.

16.2 Market Allocation

The system may allocate:

• customers;
• geographic territories;
• suppliers;
• contracts;
• delivery areas.

For example:

Institution A receives customers in Region X while Institution B receives Region Y.

If competitors intentionally use the system to divide markets, traditional cartel principles may apply.

16.3 Output Coordination

An AI system could recommend coordinated production reductions.

This could reduce:

• supply;
• capacity;
• inventory;
• investment.

The consequence could be higher prices or reduced consumer choice.

16.4 Information Asymmetry

The system operator may possess information that individual participants cannot access.

This can create:

Platform → information advantage → market power

The platform may then use its informational advantage to favour affiliated businesses.

17. Dominance and Intelligent Institutional Infrastructure

The problem is not limited to cartels.

Suppose a dominant company controls the essential coordination infrastructure.

It may:

1. deny competitors access;
2. impose discriminatory technical requirements;
3. prioritise its own services;
4. restrict interoperability;
5. degrade APIs;
6. use competitors' data;
7. impose unfair access fees.

This moves the analysis from Section 3-type coordination concerns toward abuse-of-dominance principles.

18. Essential-Facility Dimension

An intelligent institutional system can become particularly important where it functions as infrastructure.

Examples:

• national payment infrastructure;
• electricity allocation systems;
• telecommunications exchanges;
• digital identity systems;
• logistics networks;
• health-information exchanges;
• cloud infrastructure;
• dominant API platforms.

If competitors cannot realistically compete without access, refusal or discriminatory access may raise essential-facility or exclusionary-abuse concerns, depending upon the applicable jurisdiction's legal test.

19. Trade Associations as Intelligent Coordinators

Trade associations increasingly operate:

• benchmarking platforms;
• industry databases;
• compliance systems;
• certification systems;
• market intelligence platforms.

These systems can be legitimate.

But risk arises where an association collects and distributes:

current competitor-specific prices + costs + capacity + future strategy.

The association may unintentionally become the institutional hub of coordination.

20. Data Governance and Competition Law

Data governance is therefore central to IICS.

A safer information architecture can employ:

Aggregation

Instead of:

Company A = ₹100
Company B = ₹105

the system provides:

Industry average = ₹102.50

Anonymisation

Participants should not be able to identify the underlying competitor.

Time lag

Historical information may be less competitively sensitive than real-time information.

Restricted access

Competitors should not have unrestricted access to sensitive datasets.

Independent governance

A neutral administrator can reduce opportunities for strategic information exchange.

21. AI Governance Requirements

An intelligent institutional coordination system should ideally maintain:

1. purpose limitation;
2. data minimisation;
3. access controls;
4. audit trails;
5. algorithmic documentation;
6. human oversight;
7. competition-law review;
8. conflict-of-interest controls;
9. independent governance;
10. periodic antitrust audits.

22. Competition-Law Compliance Framework

A useful compliance model is:

STEP 1 — Identify participants

Who are the system's users?

STEP 2 — Identify competitive relationships

Are they actual or potential competitors?

STEP 3 — Classify information

Is information:

• public?
• aggregated?
• historical?
• confidential?
• current?
• future-oriented?

STEP 4 — Examine the algorithm

Does it:

• merely optimise?
• monitor rivals?
• predict rival conduct?
• recommend prices?
• coordinate responses?

STEP 5 — Identify the administrator

Who controls the system?

STEP 6 — Examine access

Does the administrator discriminate between users?

STEP 7 — Test competitive effects

Could the system:

• increase transparency;
• facilitate collusion;
• foreclose competitors;
• raise switching costs;
• reduce innovation?

STEP 8 — Implement safeguards

Use:

• data separation;
• aggregation;
• anonymisation;
• access controls;
• compliance monitoring.

23. Evidence in Intelligent Coordination Cases

Digital systems create unusually rich evidence.

Competition authorities may examine:

• source code;
• algorithm specifications;
• API logs;
• database architecture;
• system messages;
• audit trails;
• configuration files;
• version histories;
• communications between administrators and users;
• pricing recommendations;
• internal compliance documents;
• data-access records.

This means that competition investigations increasingly require technical as well as traditional documentary evidence.

24. The Problem of Autonomous Algorithms

One of the most difficult questions is:

Can competitors be liable where an algorithm independently learns to coordinate?

The answer depends on the applicable competition regime and facts.

A distinction should be made between:

Parallel conduct

Algorithms independently arrive at similar prices.

Facilitated coordination

Competitors knowingly use a common mechanism that facilitates alignment.

Explicit coordination

Humans intentionally configure the system to coordinate competitors.

Platform-mediated coordination

A third-party system collects competitor data and produces recommendations capable of aligning behaviour.

The mere existence of parallel algorithmic outcomes should not automatically be equated with cartelisation. Samir Agrawal illustrates the importance of this distinction in the Indian context.

25. Institutional Coordination and Innovation

There is also a pro-competitive side.

Intelligent coordination systems can:

• reduce transaction costs;
• improve supply-chain efficiency;
• facilitate interoperability;
• reduce fraud;
• improve infrastructure utilisation;
• increase consumer choice;
• improve safety;
• reduce duplication;
• facilitate technological standards.

Therefore, competition law should not treat every form of institutional coordination as suspicious.

The appropriate question is whether the system preserves independent competitive decision-making while achieving legitimate efficiencies.

26. Key Doctrinal Tests

For an IICS, the following questions are particularly important:

Test 1 — Independence

Can each competitor independently determine:

• price?
• output?
• customers?
• strategy?

Test 2 — Information

What competitively sensitive information enters the system?

Test 3 — Transparency

Can participants identify rivals' behaviour?

Test 4 — Facilitation

Does the system facilitate coordination?

Test 5 — Knowledge

Did participants know or reasonably understand the system's competitive function?

Test 6 — Intent

Was coordination deliberately designed?

Test 7 — Effects

Does the system materially affect competition?

Test 8 — Market power

Does the system operator possess substantial market power?

Test 9 — Foreclosure

Can rivals realistically compete without the system?

Test 10 — Efficiency

Are restrictions objectively connected to legitimate efficiency benefits?

27. Conceptual Model

              INTELLIGENT COORDINATION SYSTEM                         │        ┌────────────────┼────────────────┐        │                │                │      DATA            ALGORITHM        GOVERNANCE        │                │                │        ↓                ↓                ↓   Information       Processing       Rules/Access        │                │                │        └────────────────┼────────────────┘                         ↓              MARKET PARTICIPANT BEHAVIOUR                         │             ┌───────────┴───────────┐             ↓                       ↓       Legitimate coordination   Anticompetitive             │                   coordination             ↓                       ↓       Efficiency/innovation   Cartel/exclusion

28. Important Legal Distinction

The most important proposition for examination purposes is:

Competition law regulates the competitive function of institutional coordination, not the technological sophistication of the coordination mechanism.

A paper agreement, telephone call, trade-association meeting, online platform, API, AI system or common pricing algorithm can potentially perform the same economic function.

Therefore, replacing human coordination with an algorithm does not automatically remove competition-law responsibility.

At the same time, parallel algorithmic behaviour alone is not necessarily sufficient to prove collusion. The evidentiary and legal requirements remain important, as demonstrated by Eturas and Samir Agrawal.

29. Conclusion

Intelligent Institutional Coordination Systems represent a new technological layer through which competition can either be enhanced or restricted.

The principal competition-law concerns are:

• exchange of competitively sensitive information;
• hub-and-spoke coordination;
• algorithmic price alignment;
• common data infrastructure;
• intermediary facilitation;
• market allocation;
• exclusionary access;
• discriminatory interoperability;
• self-preferencing;
• dominance over essential digital infrastructure.

The six principal authorities discussed—T-Mobile Netherlands, Eturas, Dole, AC-Treuhand, Samir Agrawal and the RealPage proceedings—demonstrate the evolution from traditional human communication toward platform-mediated and algorithm-mediated coordination. The recent OECD analysis confirms that competition authorities are increasingly examining precisely this transition.

The central rule can therefore be stated simply:

An intelligent coordination system is competition-law compliant when it facilitates legitimate cooperation without undermining independent competitive decision-making; it becomes problematic when its information, algorithmic or governance architecture is used to facilitate cartelisation, exclusion or the exercise of market power.