Competition Concerns in Data Sharing Agreements

1. Introduction

Data sharing agreements are arrangements through which two or more businesses exchange, provide access to, pool, license, or jointly process data.

Data sharing can be procompetitive because it may:

improve products and services;

reduce transaction costs;

facilitate innovation;

improve fraud detection;

create interoperability;

improve logistics;

support research and development;

enable better consumer services.

However, data sharing can also create competition-law risks, particularly where competitors exchange commercially sensitive information, a dominant undertaking uses data access to exclude rivals, or several businesses pool data in a way that creates market power.

The central competition question is:

Does the data-sharing arrangement improve legitimate business activity, or does it facilitate coordination, exclusion, foreclosure, or exploitation of market power?

2. Meaning of a Data Sharing Agreement

A data-sharing agreement may involve:

customer data;

pricing data;

sales data;

inventory information;

supplier information;

production data;

transaction records;

location data;

technical data;

usage data;

advertising data;

market information;

algorithms or model-training data.

Example

Five competing retailers create a common database containing:

current prices;

future prices;

inventory;

discounts;

customer demand.

Such an arrangement can create substantial competition concerns because competitors may obtain information that reduces uncertainty about each other's future conduct.

3. Why Data Is Competition-Sensitive

Data has several economic characteristics that make it particularly important.

1. Scale

Large datasets can provide significant competitive advantages.

2. Scope

Data may be useful across multiple markets.

3. Speed

Real-time information can allow companies to react immediately to competitors.

4. Network effects

More users can generate more data, which can improve the service and attract more users.

5. Replication advantages

Some datasets may be difficult for competitors to reproduce.

6. Combination

Different datasets can be combined to produce information that individual firms could not obtain independently.

4. Main Competition Concerns

A. Exchange of Competitively Sensitive Information

This is one of the most important risks.

Competitors may exchange:

future prices;

planned price increases;

production quantities;

customer allocation;

discounts;

strategic plans;

costs;

capacity information.

Such exchanges can reduce uncertainty between competitors and facilitate coordination.

5. Data Sharing as a Facilitator of Collusion

Suppose competing firms share:

“Our price next month will be ₹100.”

Another competitor learns this information.

The information exchange can make coordinated pricing easier.

The concern becomes stronger when information is:

individualised;

current;

commercially sensitive;

non-public;

detailed;

frequent;

forward-looking.

Simplified formula

Sensitive information exchange

↓

Reduced uncertainty

↓

Easier monitoring of competitors

↓

Greater possibility of coordination

↓

Potential Article 101/antitrust concern

6. Historical vs Current Data

Not all data has the same competition significance.

Historical aggregated data

Generally presents lower coordination risks when sufficiently old and aggregated.

Current individualised data

Can create substantially greater risks.

Future strategic information

Can be particularly sensitive because it may reveal intended competitive behaviour.

Therefore:

The competitive significance of data depends not merely on the fact that data is shared, but on what data is shared, with whom, when, how precisely, and for what purpose.

7. Data Pooling

Businesses may establish a joint data pool.

Example:

Ten insurance companies contribute customer and claims information to a common database.

Potential benefits include:

fraud detection;

actuarial accuracy;

risk assessment;

improved underwriting.

Potential concerns arise if the database also provides competitors with information about:

customers;

prices;

risk strategies;

future commercial plans.

The same data pool can therefore have both efficiency benefits and competition risks.

8. Dominant Firms and Data Access

A second major concern arises when a dominant undertaking controls important data.

For example:

A dominant platform collects information from millions of businesses and then uses that information to compete against those businesses.

Possible concerns include:

leveraging;

exclusion;

discriminatory access;

self-preferencing;

refusal to provide access;

use of rivals' data;

raising rivals' costs.

The relevant legal analysis may fall under Article 102 TFEU or corresponding national competition law.

9. Data as an Entry Barrier

A large dataset can potentially become an entry barrier.

A new competitor may need:

customer histories;

transaction data;

behavioral information;

training data;

location information;

product-performance data.

If competitors cannot obtain comparable data, the incumbent may have an important advantage.

However:

Possession of a large dataset does not automatically establish dominance or an abuse.

The authority must examine the competitive importance and substitutability of the data.

10. Data Sharing and Article 101 TFEU

Article 101 may become relevant where businesses coordinate through a data-sharing agreement.

The analysis can involve:

Article 101(1)

Whether the arrangement:

has the object of restricting competition; or

produces appreciable restrictive effects.

Article 101(3)

Potential efficiencies may be relevant where the arrangement:

produces efficiencies;

gives consumers a fair share of benefits;

is indispensable to achieving those benefits; and

does not eliminate competition.

11. Data Sharing and Article 102 TFEU

Article 102 becomes relevant when the conduct involves a dominant undertaking.

Possible theories include:

refusal to provide access to important data;

discriminatory data access;

leveraging data dominance into adjacent markets;

exclusionary use of data;

tying;

self-preferencing;

discriminatory use of customer information.

12. At Least 6 Important Case Laws

1. T-Mobile Netherlands BV v Commission

Case C-8/08, CJEU, 2009

Facts

Mobile telecommunications operators participated in a meeting involving discussions concerning dealer remuneration.

Principle

The CJEU emphasized that certain exchanges of competitively sensitive information can constitute a restriction of competition by object.

Relevance to data sharing

The case demonstrates the fundamental principle that:

Information exchange between competitors can itself create competition concerns when it reduces strategic uncertainty.

A modern data-sharing platform could create similar risks if competitors receive sensitive information about each other's intended conduct.

13. 2. Eturas UAB v Lietuvos Respublikos konkurencijos taryba

Case C-74/14, CJEU, 2016

Facts

An electronic travel-booking system transmitted a message to participating travel agencies concerning a limitation on online discounts.

The system operator effectively facilitated a common pricing-related restriction.

Principle

The CJEU considered circumstances in which participation in a common electronic system could contribute to a concerted practice.

Relevance

Eturas is highly relevant to modern digital data-sharing arrangements.

It demonstrates that:

A digital platform can become relevant to competition-law analysis where it facilitates coordination between competing businesses.

The fact that information is exchanged through software rather than face-to-face communication does not remove competition concerns.

14. 3. AC-Treuhand v Commission

Case C-194/14 P, CJEU, 2015

Facts

AC-Treuhand provided services connected with cartel arrangements among producers.

Principle

The CJEU confirmed that an undertaking providing services that knowingly facilitates an anticompetitive arrangement can itself fall within Article 101 liability in appropriate circumstances.

Relevance

This is particularly significant for:

data intermediaries;

industry databases;

benchmarking platforms;

algorithm providers;

information exchanges.

A data intermediary cannot necessarily avoid competition-law scrutiny merely because it does not itself sell the competing products.

15. 4. Commission v Anic Partecipazioni

Case C-49/92 P, CJEU, 1999

Principle

The case forms part of the CJEU's jurisprudence concerning concerted practices and the concept of participation in coordinated conduct.

A concerted practice can arise from conduct that reduces the uncertainty normally associated with independent competitive behaviour.

Relevance

For data-sharing agreements, the important lesson is:

Competitors are generally expected to determine their market conduct independently.

If systematic data exchanges replace that independent decision-making process, Article 101 concerns may arise.

16. 5. United States v Container Corporation of America

393 U.S. 333 (1969)

Facts

The case concerned information exchanges among competing corrugated-container manufacturers.

Competitors exchanged information concerning prices.

Principle

The U.S. Supreme Court treated the information exchange as capable of restricting competition.

Relevance

Container Corporation is a classic authority for understanding the risks associated with competitor information exchanges.

It is particularly useful when analyzing modern data-sharing arrangements involving:

prices;

customer information;

production;

capacity;

commercial strategy.

17. 6. United States v United States Gypsum Co.

438 U.S. 422 (1978)

Facts

The case concerned information exchanges and pricing practices within the gypsum-board industry.

Principle

The Supreme Court examined the relationship between information exchanges and unlawful pricing conduct, including the significance of intent and market circumstances.

Relevance

The case demonstrates that information exchanges should be evaluated in their broader competitive context.

Not every exchange of information has the same competitive significance.

18. 7. United States v Topkins

N.D. Cal., 2015

Facts

Online sellers used pricing algorithms in connection with an agreement to coordinate prices for posters sold online.

Principle

The prosecution demonstrated that digital algorithms do not immunize coordinated pricing from antitrust law.

Relevance

The case is particularly relevant to algorithmic data sharing.

If competitors feed shared market information into algorithms that facilitate coordinated pricing, the digital nature of the arrangement does not eliminate antitrust risk.

19. 8. Hoffmann-La Roche v Commission

Case 85/76, CJEU, 1979

Although principally known for loyalty rebates, the case is important to the broader concept of dominance and market foreclosure.

Relevance

Where a dominant undertaking controls an important commercial input—including potentially strategically important information—competition analysis may examine whether its conduct strengthens market foreclosure.

The case should therefore be treated as an indirect analogy, rather than a direct data-sharing precedent.

20. Case-Law Comparison

21. Types of Data-Sharing Agreements

A. Horizontal Data Sharing

Competitors share data with each other.

Risk

Usually the most obvious cartel-related concern.

Example:

Competing airlines share future pricing and capacity plans.

B. Vertical Data Sharing

A supplier and distributor exchange information.

Example:

Manufacturer → retailer → sales data.

This can produce efficiencies but may also facilitate:

resale-price coordination;

customer allocation;

territorial restrictions;

monitoring of downstream prices.

C. Reciprocal Data Sharing

Both companies provide information to each other.

Potentially legitimate where the exchange improves:

forecasting;

logistics;

fraud prevention.

But reciprocal exchanges of highly sensitive information can facilitate coordination.

D. One-Way Data Sharing

One undertaking provides data to another.

This can raise:

dominance;

access;

discrimination;

foreclosure issues.

E. Data Pooling

Multiple businesses contribute information to a common database.

Potential benefits:

fraud prevention;

industry research;

technical standardization.

Potential risks:

market transparency;

coordination;

exclusion;

concentration of information.

22. The Role of Data Aggregation

Aggregation can reduce competition risks.

High risk

Company A's exact price for Customer X tomorrow.

Lower-risk structure

Average industry price during the previous year, covering hundreds of transactions.

Aggregation can reduce the ability to identify an individual competitor's strategy.

However, aggregation is not an automatic safe harbour.

An apparently aggregated dataset may still permit participants to reconstruct individual competitors' information.

23. Frequency of Data Sharing

Frequency matters.

Occasional historical exchange

Potentially lower coordination risk.

Daily exchange

Greater potential for monitoring.

Real-time exchange

Potentially much more significant.

For example:

Real-time competitor pricing → immediate response → easier coordination.

24. Individualized vs Aggregated Data

These are analytical indicators rather than automatic legal conclusions.

25. Data-Sharing Intermediaries

An intermediary may operate:

benchmarking platforms;

industry databases;

cloud systems;

AI platforms;

market-information services;

procurement exchanges.

The intermediary may be exposed to competition concerns if it knowingly facilitates competitor coordination.

The AC-Treuhand principle is especially relevant.

26. Algorithmic Data Sharing

Modern platforms can automatically collect and distribute data.

For example:

Competitor A

↓ data

Common algorithm

↑ data

Competitor B

The algorithm may automatically generate:

recommended prices;

inventory decisions;

customer allocations;

bidding strategies.

This creates a new form of information exchange.

The critical question is not merely whether an algorithm is involved.

It is:

Whether the underlying arrangement enables competitors to coordinate or otherwise restrict competition.

27. AI and Data-Sharing Agreements

AI increases the complexity of data-sharing arrangements because businesses may exchange:

training datasets;

model outputs;

customer behavioral data;

embeddings;

model-performance information;

demand forecasts.

Potential competition concerns include:

1. Competitor coordination

Shared AI systems may make competitor behavior more predictable.

2. Data concentration

A single AI provider may accumulate information from many competitors.

3. Exclusion

A dominant AI provider could potentially restrict rivals' access to important datasets.

4. Information asymmetry

The platform may know significantly more about participating businesses than those businesses know about one another.

28. Data Sharing and Privacy

Competition law and privacy law are different, but they can overlap.

For example:

Competition issue:
Does a data-sharing arrangement exclude competitors?

Data-protection issue:
Is the personal-data processing lawful?

A data-sharing agreement may therefore need to satisfy multiple legal regimes.

Privacy compliance does not automatically make an anticompetitive arrangement lawful.

Similarly, an antitrust concern does not automatically establish a data-protection violation.

29. Legitimate Business Justifications

Data sharing may create substantial efficiencies.

Fraud prevention

Banks share information to identify fraudulent transactions.

Supply-chain optimization

Manufacturers and distributors share inventory data.

Cybersecurity

Companies exchange threat intelligence.

Research

Companies may pool scientific datasets.

Industry standards

Businesses may exchange technical information to achieve interoperability.

Demand forecasting

Companies may share aggregated information to improve logistics.

Such efficiencies should be distinguished from exchanges designed to facilitate competitive coordination.

30. Competition Assessment Framework

A useful examination framework is:

Step 1 — Identify the participants

Are they:

competitors;

suppliers;

customers;

unrelated businesses;

joint-venture partners?

Step 2 — Identify the data

What is being shared?

price;

cost;

customer;

production;

capacity;

technical;

historical;

future strategic information.

Step 3 — Examine timing

Is the information:

historical;

current;

real-time;

forward-looking?

Step 4 — Examine aggregation

Is the data:

individualized;

anonymized;

aggregated;

easily reconstructable?

Step 5 — Examine purpose

Why is the information being shared?

Step 6 — Examine market structure

Consider:

number of competitors;

concentration;

entry barriers;

transparency;

frequency of interaction.

Step 7 — Examine effects

Does the arrangement:

facilitate coordination?

exclude competitors?

raise entry barriers?

create efficiencies?

improve consumer welfare?

31. Information Exchange Through a Platform

Suppose five competing food-delivery companies provide data to a common platform.

The platform knows:

current prices;

delivery fees;

future discounts;

driver availability;

customer demand.

If it distributes sensitive information back to competitors, the arrangement could facilitate coordination.

But if it provides only:

aggregated historical demand statistics

for legitimate forecasting, the competitive assessment may be different.

Thus:

Data sharing itself ≠ automatic antitrust violation.

The design and competitive function of the arrangement are crucial.

32. Data Sharing and Market Foreclosure

A dominant platform could potentially use data-sharing arrangements to disadvantage rivals.

For example:

A dominant marketplace requires sellers to provide detailed transaction data but prevents competing marketplaces from obtaining comparable information.

Possible effects include:

information asymmetry;

reduced rival competitiveness;

increased entry barriers;

strengthening of the incumbent's ecosystem.

This raises Article 102-type concerns where dominance and exclusionary effects are established.

33. Data Portability and Interoperability

Data sharing and data portability are related but different.

Data sharing

One company gives data to another.

Data portability

Users can move their data from one service to another.

Portability can reduce:

switching costs;

lock-in;

ecosystem dependence.

Interoperability can similarly increase competitive pressure by allowing competing services to interact.

34. Standard Safeguards for Data-Sharing Agreements

Businesses can reduce competition risks through safeguards such as:

1. Aggregation

Share only aggregated information.

2. Delay

Use historical information rather than real-time data where possible.

3. Anonymization

Remove identifying information where appropriate.

4. Independent administrator

Use an independent intermediary rather than direct competitor-to-competitor exchange.

5. Access restrictions

Limit employees who can see sensitive information.

6. Purpose limitation

Specify precisely why data is being exchanged.

7. Compliance monitoring

Regularly review the arrangement.

8. Avoid forward-looking strategic information

Particularly where competitors are involved.

35. Important Distinction: Data Is Not Automatically an Essential Facility

A company may possess a very large dataset.

That does not automatically mean:

“The dataset must be shared with competitors.”

The stringent principles from Bronner, Magill and IMS Health demonstrate that compulsory access is exceptional.

Authorities generally need to establish the relevant legal conditions rather than treating every valuable dataset as an essential facility.

36. Emerging Competition Issues

Future competition disputes may increasingly involve:

AI training-data pools;

autonomous pricing systems;

data cooperatives;

cloud data-sharing platforms;

smart-city data;

connected-vehicle data;

health-data ecosystems;

financial-data exchanges;

industrial IoT data;

biometric datasets;

real-time logistics data.

These areas may require application of established competition principles to new technological circumstances.

37. Key Case-Law Lessons

T-Mobile Netherlands

Competitors must be careful when sharing strategic information because information exchange can reduce competitive uncertainty.

Eturas

A digital platform can facilitate coordinated conduct; electronic communication does not escape Article 101.

AC-Treuhand

A third-party facilitator can potentially face competition-law liability where it knowingly contributes to an infringement.

Anic

Concerted practices can be established through conduct that substitutes coordination for genuinely independent market behaviour.

Container Corporation

Competitor information exchanges can be anticompetitive even where there is no traditional written cartel agreement.

U.S. Gypsum

The competitive significance of information exchange depends on the circumstances and relationship between information and competitive conduct.

Topkins

Algorithms and online platforms do not provide immunity from ordinary antitrust principles.

38. Short Exam-Ready Answer

Competition concerns in data-sharing agreements arise when businesses exchange or pool information in ways that facilitate coordination, reduce competitive uncertainty, create entry barriers, strengthen dominance, or exclude competitors. Horizontal exchanges of current or future prices, output, customers and strategic plans are particularly sensitive. Data-sharing arrangements can nevertheless generate legitimate efficiencies such as fraud prevention, logistics optimization, research and interoperability. Cases such as T-Mobile Netherlands, Eturas, AC-Treuhand, Anic, Container Corporation and U.S. Gypsum demonstrate the importance of information exchange in competition law, while Topkins illustrates the application of these principles to algorithmic markets. The legality of a data-sharing arrangement therefore depends on the nature of the data, timing, aggregation, purpose, market structure, competitive effects and available safeguards.