Competition Law and Governance of Commercial Knowledge Structures

Introduction

Commercial knowledge structures are the systems through which businesses create, collect, organize, exchange, analyse and control commercially relevant knowledge. They include:

• industry databases and data pools;
• benchmarking systems;
• market-intelligence platforms;
• trade associations;
• standards-development organisations;
• patent and technology pools;
• credit and business-information databases;
• algorithmic information systems;
• common procurement and distribution platforms;
• research and development networks;
• AI and machine-learning datasets;
• pricing and demand-information systems; and
• information exchanged during mergers, joint ventures and strategic collaborations.

Competition law does not generally prohibit the creation or sharing of commercial knowledge. Indeed, information can reduce information asymmetry, improve efficiency, facilitate innovation and help smaller firms compete. The central competition-law problem arises when a knowledge structure reduces competitive uncertainty, facilitates coordination, excludes rivals, or gives a dominant undertaking control over an indispensable information resource.

The OECD's 2026 review similarly identifies the basic tension: information sharing can facilitate legitimate benchmarking and cooperation, but excessive sharing can reduce the uncertainty on which competition depends.

I. Meaning of Commercial Knowledge Structures

A commercial knowledge structure can be understood as having five components:

1. Knowledge creation

Firms generate information concerning:

• prices;
• costs;
• production;
• customers;
• demand;
• capacity;
• inventories;
• technology;
• research;
• patents;
• product performance; and
• future commercial strategies.

2. Knowledge collection

Information may then be collected through:

• trade associations;
• industry databases;
• consultants;
• data brokers;
• platforms;
• common technical systems;
• market research organisations; or
• regulatory reporting systems.

3. Knowledge aggregation

Raw information can be transformed into:

• benchmarks;
• indices;
• market reports;
• forecasts;
• rankings;
• risk scores;
• algorithms;
• dashboards; and
• predictive models.

4. Knowledge distribution

The resulting information may be distributed among:

• competitors;
• suppliers;
• distributors;
• customers;
• regulators;
• investors; or
• members of an industry association.

5. Knowledge governance

Governance determines:

• who can access information;
• what information can be shared;
• when it can be shared;
• whether information is historical or prospective;
• whether it is aggregated or individualised;
• whether access is discriminatory; and
• whether the information can be used for competitive decision-making.

This final element is increasingly important because control over commercial knowledge can itself become a source of market power.

II. Competition-Law Framework

Commercial knowledge structures can implicate several major areas of competition law.

Under EU law, Article 101 TFEU is particularly important for information exchange between competitors. The European Commission's Horizontal Guidelines distinguish between information exchanges that may constitute restrictions by object and exchanges requiring an effects analysis. The Guidelines also recognise that properly structured data pools can produce pro-competitive effects where the information exchange is necessary and proportionate.

In India, the principal statutory provisions are Sections 3 and 4 of the Competition Act, 2002, supplemented by merger-control rules where knowledge structures are transferred or integrated through combinations.

III. Commercial Knowledge as a Competition Parameter

Competition does not operate only through price.

Modern firms compete through:

• information;
• innovation;
• quality;
• speed;
• reliability;
• customer intelligence;
• algorithms;
• technical standards;
• data access;
• predictive capabilities; and
• intellectual property.

Consequently, a knowledge structure can influence competition even when no price is expressly discussed.

For example, competitors exchanging information concerning their future capacity decisions may effectively communicate their intended market behaviour. Similarly, a platform controlling the industry's principal performance database may be able to disadvantage rivals without formally refusing to deal with them.

IV. Information Exchange Between Competitors

The most established competition concern is the exchange of commercially sensitive information.

Information becomes particularly problematic when it is:

• current;
• detailed;
• company-specific;
• non-public;
• commercially sensitive;
• forward-looking; and
• capable of revealing future competitive conduct.

The European Commission explains that commercially sensitive information can facilitate coordination by allowing competitors to signal desired conduct or develop mutually consistent expectations concerning market behaviour.

Particularly sensitive categories

Examples include:

1. future prices;
2. planned discounts;
3. future output;
4. customer allocation;
5. future investment;
6. capacity expansion;
7. strategic business plans;
8. individualised costs;
9. tender strategies; and
10. market-entry intentions.

By contrast, information that is:

• genuinely historical;
• aggregated;
• publicly available;
• sufficiently old; and
• incapable of revealing individual competitive strategies

will generally present a lower competition risk, although context remains decisive.

V. Six Major Case Laws

1. American Column & Lumber Co. v. United States, 257 U.S. 377 (1921)

This is one of the foundational US cases concerning commercial information structures.

A group of hardwood manufacturers established an "Open Competition Plan" through which detailed information concerning production, inventories, shipments, prices, purchasers and market conditions was collected and redistributed.

The Supreme Court found that the system operated as an unlawful restraint because the information structure facilitated coordination concerning production and prices.

Principle

A seemingly neutral information system can become anticompetitive when it operates as an instrument for coordinating competitors' commercial behaviour.

Importance

The case establishes an enduring distinction:

Information itself is not necessarily anticompetitive; the competitive function performed by the information system is critical.

2. United States v. Container Corp. of America, 393 U.S. 333 (1969)

Several corrugated-container manufacturers exchanged information concerning prices quoted to particular customers.

The Supreme Court concluded that the reciprocal exchange constituted concerted action and had the effect of stabilising prices and reducing the intensity of price competition.

Principle

An information exchange does not need to contain an explicit agreement to fix prices if the exchange itself facilitates coordination sufficiently to diminish competitive rivalry.

Significance for knowledge structures

This case demonstrates the danger of reciprocal intelligence systems:

Firm A learns Firm B's price → Firm B learns Firm A's price → uncertainty disappears → competitive reactions become predictable → price competition may weaken.

3. John Deere Ltd v Commission, Case C-7/95 P

The EU's John Deere litigation concerned an information-exchange system in the UK tractor industry.

The system enabled manufacturers to obtain detailed information concerning competitors' sales and dealer activity. The Court of Justice upheld the finding that the system could restrict competition because it substantially reduced the uncertainty normally existing between competitors.

Principle

Competition law protects a degree of strategic uncertainty between competitors.

An information system can therefore be problematic even without an express agreement concerning prices.

Key factors

The assessment included:

• market concentration;
• nature of information;
• level of aggregation;
• frequency of exchange;
• degree of transparency; and
• ability to identify individual competitors.

Modern significance

John Deere is highly relevant to:

• market-intelligence platforms;
• competitor dashboards;
• industry databases;
• automated benchmarking; and
• algorithmic pricing systems.

4. T-Mobile Netherlands BV v Raad van bestuur van de NMa, Case C-8/08

In T-Mobile Netherlands, mobile-network operators participated in a meeting in which competitively significant information was discussed.

The Court of Justice held that, in appropriate circumstances, a single meeting can be sufficient to establish a concerted practice where its object is sufficiently harmful to competition.

Principle

Competition law does not necessarily require:

• a written agreement;
• repeated meetings; or
• a formal information-sharing organisation.

A single exchange can potentially have legal consequences.

Knowledge-governance lesson

Trade associations and professional forums therefore require strong protocols governing:

• meeting agendas;
• permissible topics;
• minutes;
• information disclosure;
• participant communications; and
• departure procedures.

5. Eturas UAB and Others v Lietuvos Respublikos konkurencijos taryba, Case C-74/14

Eturas involved travel agencies using a common computerised booking system.

The system administrator communicated a message concerning restrictions on online discounts, followed by an automated restriction being implemented within the system.

The Court examined whether the circumstances could establish a concerted practice among participating undertakings.

Principle

A common technological infrastructure can become a mechanism through which competitors coordinate their commercial conduct.

Modern significance

The case is particularly relevant to:

• common platforms;
• SaaS systems;
• marketplace software;
• algorithmic pricing;
• automated discount controls;
• common APIs; and
• platform-mediated competition.

The important conceptual shift is:

Traditional cartel → human communication

versus

Digital coordination → communication + common software architecture + automated implementation.

6. HSBC Holdings plc and Others v European Commission, Case C-883/19 P

The HSBC/Euribor litigation concerned the exchange of confidential information in the Euro Interest Rate Derivatives sector and alleged manipulation concerning Euribor.

The Court of Justice addressed the legal characterisation of exchanges of confidential information and the concept of restriction by object.

Principle

The competitive significance of information must be assessed by examining:

• its content;
• its purpose;
• its economic and legal context; and
• whether it is capable of reducing strategic uncertainty.

Significance

The case illustrates how a sophisticated financial-information environment can transform apparently technical communications into competition-law issues.

7. Banco BPN v BIC Português and Others, Case C-298/22

This is a particularly important modern development.

Fourteen Portuguese banks exchanged information concerning commercial conditions and volumes in several lending markets. The Court confirmed that a standalone exchange of confidential strategic information can constitute a restriction by object when its content and context are capable of removing uncertainty concerning competitors' future conduct.

Principle

There does not necessarily have to be a separate price-fixing agreement.

The information exchange itself may constitute the competition infringement.

VI. Commercial Data Pools

Data pools create a difficult regulatory balance.

A data pool can produce substantial efficiencies by allowing firms to:

• identify fraud;
• assess credit risk;
• improve safety;
• conduct research;
• develop technical standards;
• improve forecasting;
• reduce information asymmetry; and
• create new products.

However, the same infrastructure can facilitate:

• coordinated pricing;
• market allocation;
• exclusion;
• discriminatory access;
• collective monitoring of rivals; and
• strategic transparency.

The EU Horizontal Guidelines expressly recognise that competitor data-sharing arrangements may generate pro-competitive effects where commercially sensitive information is exchanged only to the extent necessary and proportionate. Suggested safeguards include aggregation, historical data, reduced frequency and access controls.

Governance model

A compliant data pool should therefore consider:

Purpose limitation → Data minimisation → Aggregation → Access controls → Time lag → Independent administration → Audit trail

VII. Trade Associations as Knowledge Structures

Trade associations can serve legitimate purposes:

• technical education;
• industry standards;
• safety;
• statistical research;
• regulatory engagement;
• training; and
• professional development.

But they can also become platforms for:

• price discussions;
• customer allocation;
• output coordination;
• exchange of future business plans;
• collective boycotts; and
• indirect signalling.

The UK Government expressly warns that even a single meeting involving non-public commercial information can create competition-law risk.

Governance safeguards

Trade associations should therefore establish:

1. competition-law compliance policies;
2. pre-approved meeting agendas;
3. legal review of sensitive topics;
4. independent moderators;
5. written minutes;
6. prohibition on future pricing discussions;
7. controlled statistical databases;
8. aggregated reporting; and
9. immediate termination of inappropriate discussions.

VIII. Knowledge Structures and Dominance

Information exchange is not the only competition problem.

A dominant firm may control a commercially indispensable database.

Examples could include:

• a dominant credit-information database;
• a technical interoperability database;
• a critical industry standard;
• a marketplace's historical transaction dataset;
• a specialised scientific database; or
• a dominant platform's performance-information infrastructure.

The competition concern may then move from Section 3/Article 101-type coordination to abuse of dominance.

Possible theories include:

1. Refusal to provide access

A dominant firm may deny competitors access to an important information resource.

2. Discriminatory access

The dominant undertaking may provide:

• better data to itself;
• delayed access to rivals;
• inferior APIs to competitors; or
• different technical standards.

3. Excessive access conditions

Access may technically exist but be commercially unusable because of unreasonable:

• fees;
• licensing conditions;
• technical restrictions; or
• contractual limitations.

4. Self-preferencing

The owner of a knowledge platform may use its privileged access to favour its downstream products.

5. Data tying

Access to one information service may be conditioned on purchase of another product.

IX. Essential-Facility Dimension

Where a commercial knowledge infrastructure is genuinely indispensable, competition law may intersect with the essential facilities doctrine.

The general question is whether:

access to the information resource is indispensable for effective competition and whether exclusion would eliminate or seriously impede competition.

The doctrine has developed differently across jurisdictions. The US Supreme Court's Trinko decision substantially restricted compulsory-access theories in US antitrust law, whereas European competition law has retained a more developed jurisprudence concerning indispensable facilities.

Thus, not every valuable database becomes an essential facility.

The legal inquiry normally requires attention to:

• indispensability;
• replicability;
• refusal;
• competitive foreclosure;
• objective justification; and
• proportionality.

X. Standards and Knowledge Governance

Standards-development can generate enormous benefits.

Examples include:

• telecommunications standards;
• charging protocols;
• cybersecurity standards;
• payment standards;
• interoperability standards;
• technical specifications; and
• environmental measurement standards.

But standards can also become competitive bottlenecks.

Potential problems include:

Standard-setting exclusion

A rival technology may be excluded through a collectively adopted standard.

Intellectual-property leverage

A standard may incorporate patents, allowing patent holders to exercise substantial market power.

Discriminatory participation

Certain undertakings may receive privileged access to standard-setting processes.

Information asymmetry

Incumbents may possess information that entrants cannot obtain.

Consequently, competition law increasingly treats governance architecture as important, rather than looking only at the final commercial price.

XI. Mergers and Commercial Knowledge Structures

Knowledge resources can also be strategically important in merger control.

A transaction may combine:

• customer databases;
• transaction histories;
• behavioural datasets;
• intellectual property;
• technical knowledge;
• AI models;
• proprietary algorithms; and
• market intelligence.

The resulting entity may obtain capabilities that competitors cannot easily reproduce.

Gun-jumping concern

Parties to a proposed merger must also be careful about exchanging competitively sensitive information before completion.

The CCI has specifically recognised that information exchange between parties to a proposed combination can, depending on circumstances, contribute to the transaction effectively coming into effect and has emphasised proportionality and safeguards around commercially sensitive information.

Appropriate safeguards

Common mechanisms include:

• clean teams;
• restricted-access data rooms;
• anonymisation;
• aggregation;
• need-to-know access;
• independent advisers;
• information firewalls; and
• restrictions on competitively sensitive information.

XII. AI and Algorithmic Knowledge Structures

The concept becomes particularly important in AI markets.

An AI ecosystem may involve:

Data → Training dataset → Model → Algorithm → Prediction → Commercial decision

If several competitors obtain strategic information through a common AI infrastructure, the system could potentially facilitate coordination.

Potential risks include:

• algorithmic price coordination;
• common pricing engines;
• competitor monitoring;
• automated signalling;
• shared demand forecasts;
• common ranking systems;
• coordinated inventory management; and
• discriminatory access to training data.

The legal problem is not simply that AI is being used.

The relevant question is:

Does the knowledge architecture materially reduce competitive independence or exclude competing undertakings?

XIII. Commercial Knowledge and Innovation

Competition law must avoid creating excessive disincentives to legitimate knowledge sharing.

Knowledge collaboration can generate:

• technological innovation;
• lower R&D costs;
• interoperability;
• safety improvements;
• standardisation;
• better quality;
• faster product development; and
• entry opportunities for smaller firms.

The UK Government likewise recognises that many business collaborations—including R&D, production, commercialisation and information exchange—can generate consumer benefits and may be lawful depending on their circumstances.

Therefore:

Knowledge sharing ≠ automatically anticompetitive conduct.

The assessment should distinguish:

Pro-competitive knowledge infrastructure

from

coordination-enabling knowledge infrastructure.

XIV. Key Factors for Competition-Law Assessment

A regulator examining a commercial knowledge structure would typically consider:

A. Nature of information

Is it:

• public;
• confidential;
• commercially sensitive;
• historical; or
• forward-looking?

B. Degree of aggregation

Does the information identify:

• an individual undertaking;
• a customer;
• a transaction; or
• only an industry-wide trend?

C. Frequency

Is information exchanged:

• once;
• annually;
• monthly;
• daily; or
• in real time?

D. Market structure

Risk generally increases where the market has:

• few competitors;
• high entry barriers;
• homogeneous products;
• stable demand; and
• repeated interactions.

E. Purpose

Is the system designed for:

• legitimate benchmarking;
• safety;
• R&D;
• compliance;

or for:

• monitoring rivals;
• signalling;
• price alignment;
• customer allocation?

F. Governance

Who controls:

• data collection;
• aggregation;
• access;
• algorithms;
• publication; and
• enforcement?

XV. Compliance Framework for Commercial Knowledge Systems

A business or industry association should adopt a Commercial Knowledge Governance Framework.

Stage 1 — Classification

Classify information as:

• public;
• internal;
• confidential;
• commercially sensitive;
• highly strategic.

Stage 2 — Purpose limitation

Document the legitimate objective for collecting the information.

Stage 3 — Necessity

Determine whether each item of information is actually necessary.

Stage 4 — Aggregation

Where possible, use:

• industry-level;
• anonymised;
• historical; and
• statistical information.

Stage 5 — Access controls

Limit access according to:

• role;
• purpose;
• necessity;
• competition sensitivity.

Stage 6 — Independent administration

Sensitive databases should preferably be administered by an independent body rather than competitors directly.

Stage 7 — Audit

Maintain records concerning:

• data access;
• downloads;
• modifications;
• communications;
• algorithmic changes; and
• disclosure decisions.

Stage 8 — Competition review

Periodically review whether the knowledge structure has unintentionally changed from an efficiency mechanism into a coordination mechanism.

XVI. Governance Matrix

XVII. Indian Competition-Law Perspective

Under Indian competition law, commercial knowledge structures can potentially fall within several areas of the Competition Act, 2002.

Section 3

Information exchange may form part of:

• an agreement;
• arrangement;
• understanding; or
• concerted practice

that has the effect of restricting competition.

Particular attention is required where information sharing facilitates:

• price coordination;
• bid coordination;
• market allocation;
• output restriction; or
• customer allocation.

Section 4

A dominant undertaking controlling a critical knowledge infrastructure could potentially face scrutiny where its conduct constitutes:

• denial of market access;
• discriminatory conditions;
• discriminatory access;
• tying;
• unfair conditions; or
• exclusionary conduct.

Sections 5 and 6

Where commercial knowledge assets are combined through mergers, acquisitions or other combinations, competition assessment can consider whether the transaction creates or strengthens market power through control over:

• datasets;
• technology;
• IP;
• customer intelligence;
• algorithms; or
• proprietary information systems.

XVIII. Six Core Doctrinal Lessons

The principal lessons emerging from the case law are:

1. Knowledge can be a competitive instrument

American Column & Lumber demonstrates that an information system can become an instrument for restricting competition.

2. Price need not be expressly fixed

Container Corp. shows that information exchange itself may weaken price competition.

3. Competition requires strategic uncertainty

John Deere established the importance of preserving meaningful uncertainty between competitors.

4. One communication can matter

T-Mobile Netherlands demonstrates that a single meeting may be legally sufficient in appropriate circumstances.

5. Technology can become the coordination mechanism

Eturas demonstrates how a shared computer system can facilitate coordinated commercial conduct.

6. Standalone information exchange can be actionable

HSBC and the later Banco BPN/BIC Português jurisprudence demonstrate the increasing importance of analysing confidential strategic information independently of an express price-fixing agreement.

Conclusion

Governance of commercial knowledge structures is becoming a central component of modern competition law.

The traditional competition-law question was:

What price did competitors agree upon?

The modern question can be broader:

What information architecture allowed competitors to know, predict, influence, or constrain one another's competitive behaviour?

The distinction is therefore between knowledge that improves competition and knowledge structures that suppress competitive independence.

A properly designed commercial knowledge ecosystem should generally incorporate:

purpose limitation + data minimisation + aggregation + historicalisation + independent administration + access controls + information firewalls + auditability + competition-law oversight.

The seven principal authorities discussed—American Column & Lumber, Container Corp., John Deere, T-Mobile Netherlands, Eturas, HSBC, and Banco BPN/BIC Português—illustrate the evolution from traditional information exchange toward increasingly sophisticated questions concerning databases, platforms, algorithms, financial information, shared software and strategic commercial intelligence.