Competition Law and Semantic Market Structures

1. Introduction

Semantic market structures refer to markets in which competition is shaped not merely by the physical characteristics of products or services, but by the meaning, classification, interpretation, context, and relationships attached to products, data, services, search results, digital identities, and consumer preferences.

The concept is particularly important in digital and data-driven markets, where products may be differentiated through language, metadata, algorithms, taxonomies, search categories, recommendation systems, or machine-readable descriptions.

Competition law becomes relevant when a firm with market power can manipulate these semantic structures to:

• exclude competitors;
• make its own products appear more relevant;
• control how consumers discover competing products;
• restrict interoperability;
• distort product classifications;
• exploit data or metadata advantages;
• foreclose rival services through proprietary taxonomies; or
• create artificial switching costs.

Traditional competition law does not generally recognize "semantic market structure" as an independent legal category. Instead, the conduct is analysed through established doctrines such as market definition, abuse of dominance, discriminatory access, tying, refusal to deal, self-preferencing, foreclosure, interoperability restrictions, and exclusionary conduct.

2. Meaning of Semantic Market Structures

A semantic market structure exists where the competitive organization of a market depends substantially on how information is labelled, categorized, connected, ranked, interpreted, or understood.

For example, an online marketplace may classify products according to:

"premium," "budget," "eco-friendly," "recommended," "trusted," or "best value."

Although these appear to be ordinary descriptive categories, the platform may control the underlying algorithm that determines which products receive those classifications.

Consequently, control over meaning can become control over market access.

Examples

3. Semantic Structures and Market Definition

Market definition is normally concerned with identifying the products and geographic areas that constrain a firm's behaviour.

Semantic structures can complicate this analysis because products that are technically different may be commercially substitutable, while products that appear linguistically similar may not actually compete.

For example:

• "cloud storage" and "digital document management" may overlap functionally;
• "online marketplace" and "social-commerce platform" may have partially overlapping competitive functions;
• "AI assistant" and "search engine" may increasingly compete for certain queries.

Therefore, competition authorities may need to examine function, consumer purpose, data, ecosystem relationships and actual competitive constraints, rather than relying exclusively on product labels.

4. Semantic Structures and Consumer Choice

Digital platforms frequently mediate consumer decisions.

A consumer may not directly compare every available product. Instead, the platform determines:

1. what products are displayed;
2. how they are described;
3. which products are considered relevant;
4. what is recommended;
5. which products appear first; and
6. what alternatives are presented.

Consequently, semantic control can affect competition before the consumer makes a purchasing decision.

This is particularly significant where a platform acts simultaneously as:

• infrastructure provider;
• marketplace;
• search intermediary;
• data controller;
• ranking service; and
• competitor.

5. Semantic Structures and Abuse of Dominance

A dominant undertaking may potentially infringe competition law where it uses semantic control to disadvantage competitors.

Potential theories include:

A. Semantic self-preferencing

A platform could design classifications so that its own products receive favourable semantic labels.

Example:

Platform-owned products = "verified," "recommended" or "compatible"

while equivalent rival products receive inferior classifications.

B. Discriminatory classification

A platform may give competitors inferior classifications despite equivalent objective characteristics.

This can reduce:

• visibility;
• conversion rates;
• consumer trust;
• search ranking; and
• access to customers.

C. Semantic exclusion

A dominant firm might define a category in a way that excludes competing products.

For example, a platform could define "compatible devices" according to a proprietary technical standard controlled by itself.

D. Semantic tying

A platform may condition access to one service upon acceptance of its proprietary classification or metadata system.

E. Interoperability foreclosure

If competitors cannot use the dominant firm's semantic protocols, APIs, metadata standards, or classification systems, they may be unable to compete effectively.

6. Important Case Laws

There is no major body of case law expressly using "semantic market structures" as a standalone doctrine. The following cases are therefore important because they establish principles applicable to the underlying competition problems.

1. Google Search (Shopping) – European Commission

Google Search (Shopping), Case AT.39740 (2017)

The European Commission found that Google had abused its dominant position by systematically giving prominent placement to its own comparison-shopping service while applying less favourable positioning to competing comparison-shopping services.

Relevance

The case is highly relevant to semantic market structures because search relevance and ranking determine how consumers interpret competing products.

The competitive issue was not simply Google's ownership of a search engine. It concerned the way the dominant intermediary's algorithmic architecture affected the visibility of competing services.

Principle

Control over an important digital information-discovery structure can become a competition concern where it is used to favour the dominant firm's own downstream service.

2. Google Android – European Commission

Google Android, Case AT.40099 (2018)

The European Commission examined Google's conduct concerning Android devices, including restrictions involving Google Search, Chrome, Play Store licensing and manufacturers.

Relevance

The case demonstrates how control over a digital ecosystem can operate through technical and contractual structures, including default arrangements.

Semantic structures can similarly become strategically important where the dominant platform controls how applications, services or information are recognized within an ecosystem.

Principle

Control over an ecosystem can reinforce market power where restrictions make it difficult for competing services to obtain effective access to users.

3. Google AdSense – European Commission

Google Search/AdSense, Case AT.40411 (2019)

The European Commission found that Google imposed contractual restrictions on third-party websites using its search-advertising intermediation service.

Relevance

Advertising platforms operate through extensive semantic information:

• keywords;
• search terms;
• contextual classifications;
• advertiser categories; and
• user interests.

Control over those informational structures can affect which advertisers and advertisements reach consumers.

Principle

Contractual restrictions imposed by a dominant intermediary may constitute abusive exclusion where they prevent rivals from effectively competing.

4. Microsoft v Commission

Microsoft Corp. v Commission, Case T-201/04 (General Court, 2007)

The case concerned Microsoft's conduct relating to interoperability information and the Windows operating-system ecosystem.

Relevance

Interoperability is fundamentally connected to shared meaning and technical communication.

Two systems must understand:

• commands;
• protocols;
• data;
• file structures; and
• communication standards.

Where a dominant undertaking controls information necessary for interoperability, withholding or restricting that information can disadvantage competitors.

Principle

Control over interoperability information may constitute an abuse where the established legal conditions for a refusal-to-supply/interoperability theory are satisfied.

5. Bronner v Mediaprint

Oscar Bronner GmbH & Co. KG v Mediaprint, Case C-7/97

The Court of Justice considered whether a dominant newspaper-distribution system had to provide access to a competing newspaper.

Relevance

Although the case predates modern digital platforms, its essential-facilities principles are relevant to semantic infrastructures.

A semantic ecosystem may become competitively important where competitors cannot realistically operate without access to a particular infrastructure or information resource.

Principle

A refusal to provide access to infrastructure is not automatically abusive merely because the infrastructure is useful. The stringent conditions associated with compulsory access must be satisfied.

6. Slovak Telekom v Commission

Slovak Telekom a.s. and Deutsche Telekom AG v Commission, Joined Cases C-165/19 P and C-165/19 P, 2021

The case concerned access to telecommunications infrastructure and exclusionary conduct.

Relevance

Modern semantic ecosystems frequently depend upon infrastructure such as:

• APIs;
• networks;
• interoperability layers;
• data interfaces; and
• technical standards.

The case demonstrates that access conditions imposed by dominant infrastructure operators can have exclusionary consequences.

Principle

Dominant undertakings controlling infrastructure cannot use access arrangements in ways that unlawfully exclude equally efficient competitors.

7. Intel v Commission

Intel Corp. v Commission, Case C-413/14 P

The litigation concerned rebates offered by Intel and the assessment of exclusionary effects.

Relevance

Semantic structures can be accompanied by commercial incentives. For example, a platform could combine preferential classifications with:

• rebates;
• advertising advantages;
• preferred placement;
• contractual incentives.

The case illustrates the importance of examining the actual competitive effects of exclusionary conduct rather than relying solely on formal labels.

Principle

Assessment of potentially exclusionary conduct may require careful consideration of its ability to foreclose competitors and the economic circumstances surrounding the conduct.

8. Servizio Elettrico Nazionale

Servizio Elettrico Nazionale SpA v Autorità Garante della Concorrenza e del Mercato, Case C-377/20

The Court of Justice addressed the use of information acquired through a former monopoly position and its potential exploitation in competitive markets.

Relevance

This is particularly important for semantic markets because dominant firms may possess historical, behavioural, contextual or metadata advantages unavailable to rivals.

Such information may enable the dominant undertaking to classify consumers, products or market opportunities more effectively than competitors.

Principle

Information obtained through a protected or formerly monopolistic position can raise competition concerns when used to extend market power into a competitive market.

7. Common Competition Risks

7.1 Algorithmic classification

Algorithms may decide whether a product is:

• relevant;
• trustworthy;
• compatible;
• premium;
• sustainable;
• popular; or
• recommended.

If the dominant undertaking controls these classifications, competition authorities may examine whether the classification systematically disadvantages rivals.

7.2 Semantic self-preferencing

The risk is particularly significant when the platform operates downstream.

For example:

Platform → controls search meaning → classifies products → competes with those products

This creates a structural conflict.

7.3 Data advantage

Semantic systems require enormous amounts of data.

A dominant platform may have access to:

• search histories;
• click behaviour;
• purchasing patterns;
• product descriptions;
• consumer reviews;
• contextual information; and
• behavioural metadata.

This can reinforce its ability to develop superior classifications and recommendations.

8. Network Effects

Semantic systems can produce strong network effects.

More users generate more:

searches → data → classifications → better recommendations → more users.

This can create a feedback loop.

Competitors may therefore face difficulty entering even where they have a technically superior product.

Competition law may need to distinguish between:

• legitimate innovation; and
• exclusionary conduct that deliberately prevents rivals from achieving scale.

9. Semantic Lock-In

Semantic lock-in occurs when users become dependent upon a particular vocabulary, taxonomy, metadata structure or classification system.

For example:

Consumer data → Platform taxonomy → Proprietary categories → Switching costs.

A competitor entering the market may have to translate or reconstruct extensive datasets before it can provide equivalent services.

This can produce raising-rivals'-costs effects.

10. Semantic Interoperability

Interoperability requires systems to understand and process information consistently.

For example:

Platform A

"Organic certified food"

Platform B

"Certified organic product"

If the platforms cannot meaningfully interpret each other's classifications, consumers and competing suppliers may experience friction.

Competition concerns may arise where a dominant platform intentionally prevents semantic interoperability in order to protect its ecosystem.

11. Artificial Intelligence and Semantic Competition

AI systems substantially increase the importance of semantic competition.

Large language models and recommendation systems can determine:

• what products are relevant;
• which businesses are recommended;
• how competitors are described;
• whether products are considered equivalent;
• which information is presented first; and
• how consumer queries are interpreted.

A dominant AI intermediary could therefore potentially become a semantic gatekeeper.

Competition analysis may need to examine:

1. training-data advantages;
2. access to proprietary data;
3. ranking systems;
4. interoperability;
5. API access;
6. downstream integration;
7. self-preferencing;
8. exclusionary defaults; and
9. switching costs.

12. Semantic Market Structures and Merger Control

Semantic structures are also relevant to mergers.

A merger between two firms possessing complementary datasets may create a powerful semantic advantage.

For example:

Search data + purchasing data + location data + consumer reviews

could produce a much more sophisticated classification and recommendation system.

Authorities may therefore consider:

• data concentration;
• interoperability;
• innovation;
• potential competition;
• ecosystem effects;
• foreclosure; and
• access to essential data.

13. Remedies

Competition authorities may consider several remedies where semantic structures create unlawful exclusion.

Structural remedies

• divestiture;
• separation of business units;
• limits on vertical integration.

Behavioural remedies

• non-discrimination obligations;
• transparent ranking criteria;
• interoperability;
• API access;
• data portability;
• restrictions on self-preferencing.

Technical remedies

• common metadata standards;
• open APIs;
• interoperable taxonomies;
• data-transfer mechanisms;
• independent auditing of ranking systems.

14. Indian Competition-Law Perspective

In India, semantic market structures would generally be analysed under the Competition Act, 2002, particularly:

• Section 4 – abuse of dominant position;
• Section 3 – anti-competitive agreements;
• Section 5 – combinations; and
• Section 6 – regulation of combinations.

Potential Section 4 theories include:

• discriminatory conditions;
• denial of market access;
• limiting technical development;
• leveraging dominance;
• tying/bundling;
• exclusionary conduct; and
• discriminatory access to digital infrastructure.

The Competition Commission of India has increasingly dealt with digital-market issues involving platforms, algorithms, data, ranking, interoperability, app ecosystems and digital intermediation.

15. Key Legal Test

A useful analytical framework is:

Step 1 – Identify the relevant market

↓

Step 2 – Determine whether the undertaking is dominant

↓

Step 3 – Identify the semantic structure

↓

Step 4 – Determine who controls the classification/ranking/meaning

↓

Step 5 – Identify the affected competitors

↓

Step 6 – Examine foreclosure or discriminatory effects

↓

Step 7 – Examine objective justification and efficiencies

↓

Step 8 – Determine consumer and innovation effects

↓

Step 9 – Consider appropriate remedy

16. Distinguishing Legitimate Semantic Design from Anti-Competitive Conduct

Not every semantic classification is unlawful.

A platform may legitimately classify products according to:

• quality;
• safety;
• compatibility;
• consumer preferences;
• technical specifications; or
• regulatory requirements.

Competition concerns become stronger where the evidence indicates that semantic classifications are being used strategically to exclude competitors rather than to improve the product or consumer experience.

Therefore, the central question is not:

"Does the platform control meaning?"

but rather:

"Is control over meaning being used as a mechanism for unlawful exclusion or exploitation of market power?"

17. Conclusion

Semantic market structures represent an important development in modern competition analysis because competition increasingly takes place through information architecture as much as through price.

Search rankings, product taxonomies, metadata, AI classifications, recommendation systems, interoperability standards and digital identities can determine which competitors consumers see and which products they perceive as substitutes.

The cases involving Google Shopping, Google Android, Google AdSense, Microsoft, Bronner, Slovak Telekom, Intel and Servizio Elettrico Nazionale demonstrate that existing competition-law doctrines can address many of these concerns even though courts and authorities have not generally treated "semantic market structures" as a separate legal doctrine.