Competition Law and Competition Concerns in Data Trading Ecosystems

Introduction

A data trading ecosystem is a market environment in which data is collected, aggregated, enriched, licensed, exchanged, sold, brokered, or otherwise supplied between data providers, data brokers, platforms, analytics companies, advertisers, financial institutions, AI developers, and downstream businesses.

Data may include:

• consumer and transaction data;
• location and mobility data;
• search and browsing data;
• financial and credit information;
• health and insurance data;
• business and supplier information;
• advertising and audience data;
• industrial and IoT data;
• public-sector and commercially generated datasets;
• inferred or algorithmically generated profiles.

Competition law does not generally treat "data" as automatically constituting a separate market or as automatically conferring market power. The competition question is how control over data affects market entry, rivals' access, pricing, innovation, quality, and consumer choice.

In China, this issue has become particularly important in platform markets. SAMR's 2026 Internet Platform Anti-Monopoly Compliance Guidelines expressly identify the use of data, algorithms, technology and platform rules as potential sources of competition concerns and specifically discuss interruption of data sharing and control of data that may constitute an essential facility.

I. Meaning and Structure of a Data Trading Ecosystem

A typical ecosystem can be represented as:

Data Generators → Data Collectors → Data Aggregators/Brokers → Data Marketplaces → Data Users → AI/Advertising/Financial/Commercial Applications

For example:

Consumers generate purchasing and behavioural data → platform collects it → data broker aggregates it with other datasets → broker licenses an enriched dataset → advertiser uses it for targeting → competing advertiser purchases access to the same market intelligence.

Competition concerns can arise at every layer.

1. Data collection

A dominant platform may possess enormous quantities of first-party data because of its large user base.

2. Data aggregation

A firm can combine:

• transaction data;
• browsing data;
• location information;
• payment information;
• social-network information.

The resulting dataset may be substantially more valuable than each individual dataset.

3. Data brokerage

Specialised intermediaries may purchase data from numerous sources and resell or license access to downstream users.

4. Data enrichment

Data brokers may combine several datasets to produce:

• consumer profiles;
• credit scores;
• purchasing predictions;
• advertising segments;
• risk assessments;
• AI training datasets.

5. Data access

Competition problems arise where competitors cannot obtain sufficiently comparable data.

6. Data monetisation

A dominant platform may monetise data through:

• advertising;
• targeted pricing;
• commissions;
• licensing;
• analytics;
• recommendation systems;
• AI services.

II. Competition-Law Framework

Data trading ecosystems can implicate several branches of competition law.

A. Relevant Market

Possible relevant markets include:

1. data collection services;
2. data brokerage;
3. data analytics;
4. specific categories of data;
5. data-management services;
6. digital advertising;
7. data-driven financial services;
8. AI training-data markets;
9. platform services where data is an important competitive input.

The relevant market should not automatically be defined as "the data market."

The authority must examine:

• substitutability;
• data uniqueness;
• quality;
• timeliness;
• geographic scope;
• interoperability;
• switching costs;
• alternative sources;
• availability of comparable datasets.

III. Data as a Source of Market Power

Data can create market power through several mechanisms.

1. Network effects

More users generate more data.

More data improves the service.

Better service attracts more users.

More users then generate still more data.

This creates a feedback loop:

Users → Data → Better Product → More Users → More Data

2. Economies of scope

A firm possessing data from several markets may combine datasets to obtain competitive advantages unavailable to smaller rivals.

3. Economies of scale

Large platforms can spread the costs of:

• data collection;
• storage;
• processing;
• machine learning;
• cybersecurity

over enormous user bases.

4. Data uniqueness

Some datasets cannot easily be replicated.

For example, a platform may possess years of:

• individual purchasing history;
• search history;
• merchant transactions;
• location information.

5. Switching costs

Consumers may hesitate to change providers if changing platforms means losing:

• historical records;
• personalised recommendations;
• contacts;
• transaction history;
• accumulated reputation.

IV. Major Competition Concerns

1. Data Hoarding

A dominant undertaking may accumulate large quantities of commercially valuable information and deny competitors access.

The central competition question is:

Does accumulation of data merely reflect legitimate competition, or does it reinforce an existing dominant position and exclude rivals?

Data accumulation becomes particularly significant where the dataset is:

• unique;
• difficult to reproduce;
• continuously updated;
• necessary for effective competition.

V. Refusal to Provide Data

A dominant undertaking may refuse to provide data to competitors.

This may raise essential-facilities or refusal-to-deal concerns where the data is genuinely indispensable.

Under SAMR's 2026 platform compliance guidance, refusal to deal can include interruption of data sharing and, specifically, controlling data that constitutes a necessary facility and refusing to provide it on reasonable terms.

However, compulsory data access should generally be approached cautiously because forced disclosure can affect:

• investment incentives;
• privacy;
• cybersecurity;
• intellectual property;
• commercial confidentiality.

VI. Data Discrimination

A dominant data platform may provide:

• superior data access to its own subsidiary;
• inferior access to independent competitors;
• faster APIs to affiliated companies;
• preferential datasets to its own downstream services.

This can amount to discriminatory treatment when it has exclusionary effects.

Example

Suppose Platform A operates:

• a marketplace;
• an advertising business;
• a data-analytics service.

It provides its own analytics subsidiary with real-time transaction information but provides competing analytics firms with delayed or incomplete data.

The competitive concern is not simply ownership of data; it is preferential access that may disadvantage downstream rivals.

VII. Self-Preferencing Using Data

A vertically integrated platform may use data generated by independent businesses to compete against those businesses.

For example:

Marketplace → collects seller data → analyses best-selling products → launches competing private-label products

Relevant data may include:

• sales volumes;
• conversion rates;
• customer preferences;
• price elasticity;
• inventory information;
• supplier performance.

The competition concern becomes stronger where the platform has substantial market power and uses commercially sensitive information unavailable to competing sellers.

VIII. Data Tying and Bundling

A dominant undertaking may condition access to one service upon acceptance of another data-related service.

Examples include:

• mandatory data-sharing arrangements;
• compulsory analytics services;
• bundled advertising data;
• requiring merchants to use a particular data-management system;
• tying access to marketplace services to exclusive data arrangements.

Such conduct may restrict competing data providers.

IX. Exclusive Data Agreements

Exclusive data agreements can produce legitimate efficiencies.

For example, exclusivity may:

• protect investment;
• encourage data collection;
• improve data quality;
• prevent free-riding.

But exclusivity can also foreclose competitors when a dominant firm obtains exclusive access to a critical dataset.

Competition analysis therefore examines:

• duration;
• market coverage;
• exclusivity scope;
• availability of alternative data;
• market power;
• foreclosure effects.

X. Data Sharing Among Competitors

Data sharing can be either pro-competitive or anti-competitive.

Potentially pro-competitive

Sharing may improve:

• interoperability;
• fraud detection;
• cybersecurity;
• industry standards;
• research;
• logistics.

Potentially anti-competitive

Competitors sharing:

• future prices;
• customer lists;
• discounts;
• production plans;
• strategic intentions;
• capacity information

may reduce strategic uncertainty and facilitate coordination.

SAMR's 2026 guidance specifically identifies shared data pools and exchange of competitively sensitive information among competing platforms as potential sources of horizontal coordination risk.

XI. Algorithmic Coordination

Data trading can facilitate algorithmic collusion.

Suppose competing firms purchase the same real-time market dataset.

The data provider supplies:

• competitor prices;
• inventories;
• demand;
• customer behaviour.

Algorithms then automatically adjust prices.

The system can reduce uncertainty concerning competitors' behaviour and potentially facilitate coordinated outcomes.

The legal analysis should distinguish:

Independent algorithmic optimisation

from

algorithmically facilitated concerted conduct.

The latter presents substantially greater antitrust risk.

XII. Data Brokerage and Information Concentration

A major competition concern arises when one intermediary becomes the dominant supplier of commercially valuable information.

For example:

Data Provider A + Data Provider B + Data Provider C → Dominant Data Broker → Majority of downstream customers

The broker may then have incentives to:

• raise licensing prices;
• discriminate among customers;
• impose exclusivity;
• bundle datasets;
• restrict interoperability;
• deny access to competing brokers.

XIII. Mergers and Acquisitions Involving Data

Data can significantly affect merger analysis.

A transaction may combine:

• two large consumer datasets;
• an advertising platform with a data broker;
• a payment company with a consumer platform;
• an AI company with a specialised dataset provider.

Competition authorities may investigate:

Horizontal effects

Two competing data providers merge.

Vertical effects

A data supplier acquires a downstream business.

Conglomerate effects

A large platform combines datasets from unrelated markets.

Data foreclosure

The merged entity may restrict competitors' access to the combined dataset.

XIV. Chinese Competition-Law Framework

The principal framework includes:

1. Anti-Monopoly Law

Relevant areas include:

• monopoly agreements;
• abuse of dominance;
• refusal to deal;
• exclusive dealing;
• tying;
• discriminatory treatment;
• concentrations.

2. E-Commerce Law

Important for platform conduct and relationships with platform merchants.

3. Anti-Unfair Competition Law

Particularly relevant where unauthorised technological extraction, copying or misuse of commercially valuable data affects competitive interests.

4. Data-related legislation

Data-related competition analysis also interacts with:

• Data Security Law;
• Personal Information Protection Law;
• Cybersecurity Law.

These regimes do not replace competition law but may affect whether particular forms of data access or sharing are legally permissible.

XV. Important Case Laws

1. IMS Health GmbH & Co. OHG v NDC Health GmbH & Co KG

European Court of Justice, C-418/01

This is a foundational case for competition law involving commercially valuable information.

The dispute concerned IMS Health's control over a pharmaceutical-sales data structure used by competitors.

The Court developed stringent conditions for compelling access to an intellectual-property-related resource.

Competition significance

The case demonstrates that:

• possession of valuable information does not automatically create an obligation to license;
• indispensability is important;
• refusal must be capable of excluding effective competition;
• access must not be imposed merely because competitors prefer the incumbent's resource.

Relevance to data trading

The IMS framework is highly relevant where a dominant undertaking controls a dataset that competitors claim is indispensable.

2. Bronner v Mediaprint

ECJ, C-7/97

Although not a pure data case, Bronner is fundamental to refusal-to-deal analysis.

The Court emphasised the demanding nature of the conditions under which a dominant undertaking can be required to provide access to an infrastructure or resource.

Data relevance

A data platform cannot be required to share every commercially valuable dataset merely because rivals would benefit from access.

The claimant normally needs to demonstrate something approaching genuine indispensability.

3. Slovak Telekom v European Commission

CJEU, Joined Cases C-165/19 P and C-166/19 P

The case concerned access to infrastructure and exclusionary conduct.

Data-ecosystem relevance

The case is useful for understanding the broader distinction between:

• legitimate control over an input; and
• strategic use of control over an important input to exclude competitors.

The same analytical problem arises where the controlled input is data rather than telecommunications infrastructure.

4. Google Shopping

European Commission / General Court, Google Search (Shopping)

The case concerned Google's treatment of its comparison-shopping service and competing services.

Data relevance

The case illustrates how a platform controlling a major gateway can use:

• ranking;
• visibility;
• algorithms;
• user information;
• platform infrastructure

to influence downstream competition.

The lesson for data trading ecosystems is that competition concerns may arise from the combination of data and platform control, rather than from ownership of data alone.

5. Bundeskartellamt v Meta Platforms — Facebook Data Case

German Federal Cartel Office, 2019; Federal Court of Justice, 2020/2021

The German proceedings concerned Meta's combination of user information from different sources and the relationship between data practices and market power.

Competition significance

The case is particularly important because it demonstrated that data collection and combination can become relevant to abuse-of-dominance analysis.

It raised questions concerning:

• cross-service data combination;
• user data accumulation;
• privacy-related conditions;
• market power;
• exploitation of users;
• competition between social-network services.

Data-trading significance

A platform can potentially strengthen its competitive position by combining datasets from multiple services.

6. FTC v Facebook / Meta

United States Federal Trade Commission

The US litigation concerning Facebook/Meta involved competition issues surrounding the company's position in personal social networking and its acquisitions.

Relevance to data ecosystems

The case illustrates the importance of data-driven network effects and acquisitions where:

more users → more data → better targeting/service → more users.

It also demonstrates why competition authorities scrutinise acquisitions involving data-rich digital platforms.

7. MEO v Autoridade da Concorrência

Court of Justice of the European Union, C-525/16

This case concerned discriminatory pricing by a dominant undertaking.

Data ecosystem relevance

The principles are relevant where a data platform provides different customers with different access terms.

For example:

• Customer A receives real-time data;
• Customer B receives delayed data;
• Customer C pays substantially more;
• affiliated Customer D receives preferential access.

The key question is whether the differential treatment produces competitive disadvantage and whether there is objective justification.

8. Alibaba Group — “Choose One from Two”

SAMR, 2021

This is one of China's most important platform competition cases.

SAMR found that Alibaba had a dominant position in China's online retail platform services market and had required merchants to choose between Alibaba and competing platforms.

The conduct was supported through platform rules, market power, data and algorithmic mechanisms. SAMR treated the conduct as abuse of dominance through exclusive dealing and imposed a fine of RMB 18.228 billion.

Data-trading significance

The case demonstrates that:

• data;
• algorithms;
• platform rules;
• merchant relationships

can operate together to reinforce exclusionary conduct.

9. Meituan — “Choose One from Two”

SAMR, 2021

The Meituan case similarly concerned exclusive dealing in the platform economy.

Relevance to data ecosystems

Platform exclusivity can prevent rivals from obtaining:

• merchant data;
• consumer transaction data;
• ordering information;
• behavioural information.

Thus, an apparently contractual exclusivity arrangement can have important data-accumulation consequences.

10. CNKI — Abuse of Dominance

SAMR, 2022

China's investigation of CNKI is significant because the company operated an important academic-information platform.

The case involved concerns relating to:

• exclusive cooperation;
• database resources;
• licensing;
• pricing;
• access to academic information.

Data-trading significance

It demonstrates that database control can have competition implications even outside traditional consumer-facing platforms.

11. Tencent–China Music / Kugou–Kuwo and Related Digital-Content Cases

Chinese platform competition enforcement involving Tencent and digital-content ecosystems demonstrates the importance of control over valuable digital resources, licensing relationships and platform access.

Data-ecosystem lesson

Digital-content markets can generate competition concerns where control over:

• user data;
• content information;
• licensing;
• recommendation systems;
• platform traffic

creates barriers to rival entry.

12. Tencent–Himalaya Concentration Case

SAMR, 2026

SAMR conditionally approved Tencent's acquisition of a stake in Himalaya after concluding that the transaction could have effects restricting competition in China's online-audio and online-music platform markets.

Relevance

The case illustrates the importance of examining acquisitions involving large digital ecosystems where:

• user bases overlap;
• datasets can be combined;
• platform ecosystems interact;
• access to downstream markets may change after consolidation.

It is therefore relevant to the data-concentration dimension of merger control.

XVI. Data Portability as a Competition Remedy

One potential remedy is data portability.

A customer could transfer:

• transaction history;
• account history;
• usage data;
• reputation information;
• relevant business data

from one provider to another.

This can reduce:

• switching costs;
• lock-in;
• entry barriers.

However, portability must be carefully designed because not every dataset can safely or legally be transferred.

XVII. Interoperability

Interoperability can be particularly important where several platforms operate interconnected services.

Competition authorities may examine:

• APIs;
• data formats;
• technical interfaces;
• authentication;
• data-transfer protocols.

A dominant platform that unnecessarily prevents interoperability may make it difficult for competitors to enter.

SAMR's 2026 compliance guidance specifically identifies closing interfaces and interrupting data sharing among the forms of conduct that may create refusal-to-deal concerns for dominant platforms.

XVIII. Data as an Essential Facility

The essential-facility argument is strongest where the dataset is:

1. controlled by a dominant undertaking;
2. genuinely indispensable;
3. difficult or impossible to reproduce;
4. unavailable through reasonable alternatives;
5. capable of supporting competition downstream;
6. capable of being supplied without undermining legitimate interests.

Example

Suppose one company possesses the only comprehensive dataset recording millions of real-time transactions in a specialised market.

A competing analytics provider requests access.

The authority may examine:

Is the dataset genuinely indispensable, or can the rival collect sufficiently equivalent information elsewhere?

This distinction is crucial.

XIX. Predatory Data Acquisition

A firm may theoretically acquire data below sustainable commercial value to deprive competitors of access or prevent the emergence of alternative data suppliers.

Possible strategies include:

• acquiring multiple small data companies;
• signing extensive exclusivity contracts;
• purchasing datasets merely to prevent rival access;
• absorbing emerging data brokers.

This may raise concerns under both:

• abuse-of-dominance rules; and
• merger control.

XX. Data-Marketplace Self-Preferencing

Consider:

Data Marketplace → hosts 50 suppliers → operates its own analytics service

If the marketplace uses confidential supplier information to improve its own competing analytics product, it may gain an advantage unavailable to independent competitors.

The competition inquiry should consider:

• market position;
• nature of the information;
• exclusivity;
• downstream competition;
• foreclosure;
• efficiency justifications.

XXI. Data Quality as a Competition Parameter

Competition is not limited to price.

Data markets compete on:

• accuracy;
• completeness;
• freshness;
• reliability;
• granularity;
• geographic coverage;
• interoperability;
• security.

A dominant firm could theoretically weaken competition by:

• degrading rival access;
• supplying incomplete datasets;
• delaying updates;
• restricting data formats.

Thus, quality competition must be incorporated into the analysis.

XXII. Privacy and Competition

Privacy can become a competition parameter.

Consumers may value:

• less data collection;
• greater control;
• confidentiality;
• transparency.

A dominant platform that imposes substantially worse privacy conditions may potentially be engaging in a form of non-price competitive harm.

However, competition authorities must distinguish:

privacy violations as such

from

competition harm arising from the exploitation of market power.

The two legal inquiries can overlap but are not identical.

XXIII. Data Trading and Consumer Exploitation

A dominant data intermediary could potentially:

• charge excessive data-access prices;
• impose unfair licensing terms;
• demand extensive data rights;
• restrict customers' ability to reuse their data.

Competition analysis may therefore involve both:

Exclusionary abuse

Harming competitors.

Exploitative abuse

Imposing unfair conditions on trading partners or users.

XXIV. Data Trading and AI

The development of generative AI significantly increases the importance of data markets.

AI developers compete for:

• training datasets;
• specialised industry datasets;
• labelled datasets;
• real-time data;
• proprietary consumer data;
• copyrighted or licensed information.

Competition concerns may arise where a dominant platform:

controls a critical dataset + controls distribution + operates an AI service.

It may then have incentives to reserve data for its own AI system or offer competitors data only on discriminatory terms.

XXV. Data Trading and Merger Remedies

Potential remedies include:

Structural remedies

• divestiture of a data business;
• separation of datasets;
• prohibition of certain acquisitions.

Behavioural remedies

• non-discrimination;
• data-access obligations;
• interoperability;
• API access;
• licensing commitments;
• prohibition of exclusive data arrangements.

Data remedies

• portability;
• data-sharing protocols;
• independent data-access mechanisms;
• separation of competitively sensitive information.

XXVI. Compliance Checklist for Businesses

A company operating a data trading ecosystem should assess:

XXVII. Key Legal Principles Emerging from the Case Law

The cases collectively demonstrate several important principles.

Principle 1 — Data ownership is not automatically market dominance

Possession of a large dataset does not by itself establish dominance.

Principle 2 — Indispensability matters

Refusal-to-share claims require careful examination of whether alternative datasets genuinely exist.

Principle 3 — Data can reinforce network effects

Large user bases can generate data advantages that make entry progressively more difficult.

Principle 4 — Exclusive access can foreclose competitors

Exclusivity becomes particularly problematic when the dataset is difficult to reproduce.

Principle 5 — Data and algorithms should be analysed together

A dataset may have limited competitive significance in isolation but become highly powerful when combined with algorithms, platform infrastructure and network effects.

Principle 6 — Data can be both an input and a competitive advantage

The same dataset may simultaneously constitute:

• an economic input;
• a source of market power;
• a strategic asset;
• a competitive advantage.

Principle 7 — Competition law does not require automatic data sharing

Compulsory access must be balanced against:

• investment incentives;
• confidentiality;
• privacy;
• security;
• intellectual-property interests.

Conclusion

Data trading ecosystems create a distinctive competition-law problem because control over data can simultaneously affect market structure, entry, innovation, pricing, advertising, AI development and consumer choice.

The principal competition concerns are:

1. data monopolisation;
2. exclusive data access;
3. refusal to provide indispensable datasets;
4. discriminatory data access;
5. data-driven self-preferencing;
6. data tying and bundling;
7. anti-competitive data sharing;
8. algorithmic coordination;
9. data-driven merger concentration;
10. data portability and interoperability barriers;
11. misuse of commercially sensitive information;
12. foreclosure of competing data brokers and downstream businesses.

The modern approach is therefore not simply “Who owns the data?” but rather:

Who controls the data, how difficult is it for competitors to reproduce, how is that control used, and what effect does that use have on competition?

For China specifically, recent SAMR enforcement shows an increasingly technology-focused approach. Its 2026 platform guidance expressly addresses data sharing, APIs, algorithms, competitively sensitive information and essential data, while recent enforcement against Alibaba and the 2026 Tencent–Himalaya merger review demonstrate the continuing importance of data and digital ecosystems in Chinese competition law.