Competition Law and Intelligent Market Evolution and Antitrust

1. Introduction

Intelligent market evolution refers to the transformation of markets through artificial intelligence, algorithms, big-data analytics, automated decision-making, platform ecosystems, predictive technologies, digital intermediaries, network effects, and continuously adapting business models.

Traditional competition law generally examines market power through relatively static concepts—relevant market, market share, barriers to entry, pricing, output, and consumer welfare. Intelligent markets create a more dynamic environment in which:

• algorithms continuously alter prices and rankings;
• AI systems learn from competitors and consumers;
• platforms become more valuable as users join them;
• data accumulates as a competitive asset;
• ecosystems expand into adjacent markets;
• switching costs and interoperability can determine competitive conditions;
• an apparently competitive market can rapidly become concentrated;
• technological innovation itself can become a mechanism of exclusion.

Thus, antitrust must examine not merely who possesses market power today, but how technological feedback mechanisms may create, preserve, transfer, or amplify market power over time.

The Chinese platform-economy experience illustrates this dynamic particularly clearly: the 2021 SAMR action against Alibaba concerned exclusive dealing by a major platform, while subsequent scholarship has treated the case as an important example of how enforcement can affect competitive conditions in an evolving digital ecosystem.

2. Meaning of Intelligent Market Evolution

An intelligent market can be understood as a market in which data and automated computational systems materially influence competitive behaviour.

Its evolution can occur through five principal mechanisms:

A. Data-driven evolution

A platform collects:

• consumer data;
• transaction data;
• search histories;
• location information;
• purchasing patterns;
• supplier information;
• competitor information.

The larger the platform becomes, the more data it obtains. That data improves its algorithms, which may attract more users, producing still more data.

This creates a potential:

Data → Algorithm → Better service → More users → More data

feedback loop.

B. Algorithmic evolution

Algorithms can automatically modify:

• prices;
• search rankings;
• advertising;
• product recommendations;
• commissions;
• discounts;
• inventory allocation;
• access conditions.

Consequently, market behaviour can change much faster than traditional competition-law investigations can observe.

C. Network-effect evolution

Digital platforms frequently exhibit direct or indirect network effects.

For example:

More consumers → more sellers → greater platform attractiveness → more consumers.

Once a platform reaches sufficient scale, rivals may face difficulty overcoming the incumbent's network advantages even if their underlying technology is competitive.

D. Ecosystem evolution

A technology company may begin in one market and progressively expand into:

search → advertising → payments → cloud → AI → devices → content → financial services.

The competition issue therefore becomes broader than traditional single-market dominance.

The European Commission's recent Google DMA enforcement illustrates this ecosystem-oriented approach: in July 2026, the Commission found Google non-compliant with DMA obligations concerning self-preferencing in Search and restrictions on steering users toward alternative purchasing channels in Google Play.

E. Predictive evolution

AI allows undertakings to anticipate:

• consumer demand;
• competitor behaviour;
• price movements;
• supply shortages;
• customer switching;
• advertising responses.

This creates a new antitrust question:

Can competition law regulate competitive harm produced by systems that predict and automatically respond to market behaviour?

3. Competition-Law Problems Created by Intelligent Market Evolution

3.1 Dynamic Market Definition

Traditional market definition may become inadequate when technology rapidly changes substitution patterns.

For example, consumers may substitute between:

• physical retail and e-commerce;
• search engines and AI assistants;
• hotels and short-term accommodation platforms;
• banks and fintech platforms;
• traditional advertising and algorithmic advertising.

The relevant market may therefore evolve while the investigation itself is taking place.

Competition authorities should consider:

1. current substitution;
2. potential substitution;
3. innovation competition;
4. technological entry;
5. ecosystem expansion;
6. multi-homing;
7. switching costs;
8. data advantages.

4. Dynamic Market Power

Market share alone may not accurately reveal power in intelligent markets.

A platform with relatively modest current market share could possess substantial future competitive significance because it controls:

• an important dataset;
• a critical API;
• an operating system;
• a cloud infrastructure;
• a payment network;
• an AI model;
• an app marketplace;
• a distribution channel.

Conversely, a company with a high market share may face significant competitive pressure from rapid technological innovation.

Therefore:

Static market share + dynamic competitive constraints = better assessment of intelligent-market power.

5. Network Effects and Market Tipping

Network effects can create a self-reinforcing competitive structure.

Typical cycle

Users increase
↓
More data generated
↓
Algorithms improve
↓
Service quality increases
↓
More users join
↓
Competitors lose scale
↓
Market tips toward concentration

The problem is not network effects themselves. Network effects can generate substantial efficiencies.

The competition concern arises when a dominant undertaking uses those effects to:

• exclude rivals;
• impose exclusivity;
• prevent interoperability;
• restrict data portability;
• foreclose distribution;
• tie complementary products;
• discriminate against competing services.

6. Six Important Case Laws

Case 1: Google Search (Shopping) — European Union

Google Search (Shopping), Google and Alphabet v European Commission

This is one of the most important cases concerning competition in algorithmically organised markets.

The European Commission found that Google had given preferential treatment to its own comparison-shopping service in general search results while demoting competing comparison-shopping services.

The significance for intelligent market evolution is that ranking itself became a competitive instrument.

Competition principle

A platform controlling an important algorithmic gateway cannot necessarily use that gateway to systematically disadvantage competing services.

Relevance

The case demonstrates that market power can arise not merely from price control but from control over:

• visibility;
• ranking;
• traffic;
• consumer attention;
• algorithmic access.

The issue has subsequently acquired greater significance under the EU's DMA framework, including the Commission's 2026 Google Search self-preferencing finding.

7. Case 2: Google Android — European Union

Google Android, Case AT.40099

The European Commission examined Google's conduct concerning the Android mobile ecosystem, including tying and contractual arrangements involving:

• Google Search;
• Google Play Store;
• browser applications;
• Android device manufacturers.

Competition principle

An undertaking controlling a technological ecosystem may be able to use dominance in one layer to reinforce its position in another.

Intelligent-market significance

The case illustrates ecosystem leverage:

Operating system → App store → Search → Browser → Data → Advertising

The competitive concern therefore moves beyond a single product market.

Broader lesson

Competition law must examine whether technological integration:

• creates efficiencies;
• improves interoperability;
• reduces costs;

or instead:

• forecloses rivals;
• raises entry barriers;
• reinforces dominance.

8. Case 3: Microsoft — European Union

Microsoft Corp. v Commission / Microsoft interoperability cases

Microsoft's conduct concerning interoperability and integration of software products became a foundational example of competition law dealing with technology ecosystems.

The case concerned issues including:

• interoperability;
• tying;
• technological integration;
• access to information necessary for competing products.

Intelligent-market principle

Control over an important technological interface can provide an undertaking with structural power over downstream competition.

The importance today extends to:

• APIs;
• cloud platforms;
• operating systems;
• AI interfaces;
• data interoperability.

Thus:

Control of infrastructure can become control of competitive opportunity.

9. Case 4: Amazon Marketplace — European Union

The European Commission's Amazon investigations examined the use of marketplace data and the relationship between Amazon's platform and third-party sellers.

The central competition concern was that a platform simultaneously acting as:

1. marketplace operator, and
2. competing retailer,

may possess information about rival sellers that is unavailable to those sellers themselves.

Intelligent-market significance

This illustrates data asymmetry.

Amazon's platform may observe:

• sales;
• prices;
• demand;
• inventory;
• seller performance;
• consumer behaviour.

A platform's information advantage can potentially become a competitive advantage over businesses that depend upon the platform.

General principle

The more a platform acts simultaneously as:

Infrastructure + intermediary + competitor

the greater the importance of examining conflicts between platform governance and competition.

10. Case 5: Booking.com — EU/Germany

Booking.com BV v 25hours Hotel Company Berlin GmbH and Others, Case C-264/23

The Court of Justice of the European Union addressed price-parity clauses used by Booking.com.

The Court held in 2024 that the relevant parity clauses could not simply be treated as ancillary restraints outside Article 101(1) TFEU.

Intelligent-market significance

Online platforms can influence competition through contractual architecture rather than direct pricing.

A platform may regulate:

• prices;
• ranking;
• access;
• commissions;
• visibility;
• parity;
• consumer traffic.

The case therefore demonstrates that platform rules themselves can shape market evolution.

The EU subsequently applied the Digital Markets Act to Booking.com, including prohibitions affecting parity practices and obligations concerning data access.

11. Case 6: SAMR v Alibaba — China

State Administration for Market Regulation v Alibaba, 2021

This is a particularly important Chinese digital-platform case.

SAMR concluded that Alibaba had abused its dominant position through an exclusivity arrangement commonly described as "choose one from two," restricting merchants from operating through competing platforms.

SAMR imposed a fine of approximately RMB 18.228 billion, equivalent to 4% of Alibaba's 2019 domestic sales.

Intelligent-market significance

The case demonstrates how platform dominance can evolve through:

• network effects;
• merchant dependence;
• accumulated consumer traffic;
• data advantages;
• ecosystem scale;
• exclusivity.

The case is particularly relevant to the concept of intelligent market evolution because platform dominance can become self-reinforcing:

More consumers → more merchants → more transactions → more data → stronger platform → greater merchant dependence.

Recent empirical research examining the case found substantial effects on Alibaba's profitability and market response following enforcement, although those findings are economic research rather than judicial conclusions.

12. Case 7: Google Search and Self-Preferencing — Modern EU Digital Regulation

A further important development is the transition from traditional ex-post antitrust toward ex-ante regulation of structurally important digital platforms.

The EU Digital Markets Act requires designated gatekeepers to comply with specific obligations concerning platform conduct.

The 2026 Google enforcement action concerning Search and Google Play demonstrates how competition regulation is increasingly addressing algorithmic and ecosystem behaviour directly, rather than waiting for conventional dominance litigation to establish every element of an Article 102-type abuse.

This represents an important evolution:

Traditional antitrust → digital antitrust → ecosystem regulation → ex-ante platform regulation.

13. Case 8: Facebook/Meta — Germany

Bundeskartellamt v Facebook/Meta

The German competition authority's Facebook case concerned the relationship between social-network dominance and the collection/use of data from different sources.

Its importance lies in recognising that competition problems in digital markets may involve:

• data concentration;
• user lock-in;
• privacy-related conditions;
• cross-platform data aggregation;
• network effects.

Germany subsequently developed Section 19a GWB, allowing special intervention against undertakings of paramount significance for competition across markets. The regime has been used in relation to major technology companies including Alphabet/Google, Amazon and Meta.

14. Algorithmic Pricing and Tacit Coordination

Intelligent markets create an especially difficult problem where algorithms independently observe competitors' prices and adjust automatically.

Consider:

Firm A algorithm → observes Firm B → raises price

Firm B algorithm → observes Firm A → raises price

The result may be higher prices without a conventional human agreement.

The competition-law question becomes:

When does intelligent adaptation constitute legitimate independent conduct, and when does it become prohibited coordination?

Relevant indicators can include:

• common algorithm providers;
• exchange of competitively sensitive information;
• explicit instructions to coordinate;
• predictable algorithmic responses;
• communication between competitors;
• algorithm design deliberately facilitating coordination.

Mere parallel pricing is not automatically proof of a cartel.

15. AI and Collusion

AI systems may intensify traditional coordination risks because algorithms can:

• monitor markets continuously;
• detect competitor movements instantly;
• experiment with pricing;
• predict competitor reactions;
• punish deviations;
• optimise prices simultaneously.

This raises a fundamental distinction:

Traditional cartel

Human agreement → coordinated conduct

Algorithmic cartel

Human agreement → algorithmic implementation

Autonomous coordination

Algorithms independently learn mutually accommodating behaviour

The third category creates the most difficult doctrinal problem because conventional concepts of "agreement" and "concerted practice" may not map neatly onto autonomous systems.

16. Data as a Competitive Asset

In intelligent markets, data may function similarly to traditional strategic assets.

Competition authorities may examine:

Thus, competition law increasingly needs to ask:

Who controls the data necessary to compete?

17. AI and Essential-Facility-Type Problems

An AI ecosystem may depend upon access to:

• computing capacity;
• cloud infrastructure;
• training datasets;
• foundation models;
• APIs;
• application stores;
• payment systems;
• technical standards.

If a dominant undertaking controls an indispensable input, refusal or discriminatory access may create competition concerns.

However, not every commercially important input constitutes an essential facility.

Authorities normally need to distinguish:

1. genuine indispensability;
2. ordinary commercial dependency;
3. technological superiority;
4. legitimate IP protection;
5. exclusionary foreclosure.

18. Innovation Competition

Intelligent markets require competition law to protect not merely present price competition but also future innovation competition.

A dominant firm could theoretically suppress competition by:

• acquiring an emerging technology;
• denying interoperability;
• copying and degrading a rival's functionality;
• restricting access to essential data;
• tying new AI services to established products;
• controlling distribution channels.

Therefore, merger analysis increasingly needs to consider:

Who could become a significant competitor tomorrow?

This is particularly important in technology markets where today's small undertaking may possess tomorrow's disruptive technology.

19. Killer Acquisitions and Intelligent Markets

Traditional merger analysis focuses on current turnover and market shares.

Intelligent markets complicate this because a target may have:

• low current revenues;
• valuable data;
• advanced AI;
• strong user growth;
• significant R&D;
• important patents;
• technological capabilities.

Consequently, a transaction may eliminate a potential innovation competitor before that competitor becomes a substantial market participant.

The Microsoft/Activision and Booking/eTraveli decisions illustrate the broader difficulty of assessing ecosystem effects dynamically: comparative analysis of those cases identifies a tension between traditional static indicators and forward-looking ecosystem theories of harm.

20. Consumer Welfare in Intelligent Markets

Consumer welfare should not be reduced exclusively to price.

Digital consumers may receive services at a monetary price of zero.

Competition analysis may therefore examine:

• quality;
• privacy;
• security;
• choice;
• innovation;
• speed;
• interoperability;
• advertising intensity;
• data extraction;
• switching costs.

A platform can potentially harm competition even while maintaining a zero monetary price.

21. Switching Costs and Lock-In

Intelligent ecosystems may create substantial switching costs.

Examples include:

• loss of historical data;
• incompatible formats;
• loss of social connections;
• loss of accumulated reputation;
• subscription bundles;
• device dependence;
• proprietary APIs;
• loyalty programmes;
• ecosystem-specific applications.

This creates:

User lock-in → reduced switching → stronger incumbent position → greater data accumulation → further lock-in.

Competition law should therefore examine whether switching barriers are natural consequences of innovation or deliberately constructed exclusionary mechanisms.

22. Interoperability as a Competition Remedy

Where market power is reinforced by technological isolation, interoperability can become an important remedy.

Possible measures include:

• API access;
• data portability;
• technical interoperability;
• non-discriminatory access;
• switching tools;
• standardisation;
• separation of platform and downstream functions.

But interoperability remedies must balance competition with:

• cybersecurity;
• privacy;
• intellectual-property rights;
• system integrity;
• technical feasibility.

23. Self-Preferencing

Self-preferencing occurs when a platform gives its own product or service preferential treatment over competing products.

Examples include:

• search ranking;
• recommendation systems;
• app-store placement;
• marketplace rankings;
• advertising placement;
• default settings.

The modern Google enforcement record demonstrates why algorithmic ranking has become a central competition issue.

The legal inquiry should examine:

1. dominance or gatekeeper status;
2. control over the relevant gateway;
3. discriminatory treatment;
4. foreclosure;
5. effects on rivals;
6. consumer effects;
7. objective justification;
8. efficiencies.

24. Intelligent Market Evolution and Abuse of Dominance

Potential forms of abuse include:

1. Algorithmic exclusion

Using algorithms to systematically disadvantage rivals.

2. Data leveraging

Using information obtained in one market to compete unfairly in another.

3. Self-preferencing

Promoting proprietary services.

4. Algorithmic discrimination

Applying different conditions to competing businesses.

5. Exclusivity

Preventing suppliers or merchants from using rival platforms.

6. Tying

Making access to one service conditional upon use of another.

7. Predatory algorithms

Using automated systems to sustain below-cost strategies designed to exclude rivals.

8. Interoperability restrictions

Preventing rivals from technically connecting with the dominant ecosystem.

25. Intelligent Markets and Merger Control

Merger authorities should consider:

Static factors

• market shares;
• concentration;
• entry barriers;
• closeness of competition.

Dynamic factors

• innovation pipelines;
• technological capabilities;
• data assets;
• user growth;
• network effects;
• interoperability;
• ecosystem expansion;
• potential competition.

Thus:

Competition today + innovation tomorrow = dynamic merger assessment.

26. Role of Market Investigation

Competition authorities increasingly need technical expertise in:

• machine learning;
• algorithm auditing;
• data science;
• econometrics;
• network analysis;
• software architecture;
• cybersecurity;
• cloud computing.

Traditional documentary evidence may be insufficient.

Authorities may need to examine:

• source-code documentation;
• algorithmic logs;
• model-training information;
• A/B testing;
• ranking criteria;
• API records;
• pricing histories;
• internal communications.

27. Evidentiary Problems

Intelligent markets create several evidentiary challenges.

Black-box problem

The decision-maker may not understand precisely why an algorithm produced a result.

Explainability problem

A company may claim that the outcome was produced automatically.

Attribution problem

Who is legally responsible?

• developer?
• platform?
• management?
• algorithm?
• third-party vendor?

Causation problem

It may be difficult to prove that algorithmic conduct caused:

• exclusion;
• higher prices;
• reduced innovation;
• reduced consumer choice.

28. Important Legal Test

A useful analytical framework is:

STEP 1 — Define the market

Identify:

• product;
• geography;
• technology;
• user group;
• platform side.

STEP 2 — Identify intelligent infrastructure

Determine whether the undertaking controls:

• data;
• algorithms;
• cloud;
• APIs;
• operating systems;
• platforms;
• distribution.

STEP 3 — Determine market power

Examine:

• market share;
• network effects;
• switching costs;
• entry barriers;
• data advantages;
• ecosystem strength.

STEP 4 — Identify conduct

Ask whether the undertaking uses:

• exclusion;
• tying;
• exclusivity;
• self-preferencing;
• discriminatory access;
• algorithmic coordination.

STEP 5 — Examine effects

Assess:

• foreclosure;
• prices;
• quality;
• innovation;
• consumer choice;
• competitors;
• market entry.

STEP 6 — Consider efficiencies

Examine:

• innovation;
• quality improvement;
• security;
• integration;
• reduced transaction costs.

STEP 7 — Apply remedy

Potential remedies include:

• behavioural commitments;
• interoperability;
• data portability;
• non-discrimination;
• access obligations;
• structural separation;
• fines;
• merger remedies.

29. Six Core Case-Law Principles — Summary

30. Emerging Doctrine: From Static Antitrust to Dynamic Antitrust

The development can be represented as:

Traditional Competition Law
↓
Market definition
↓
Market share
↓
Dominance
↓
Conduct
↓
Effects

Digital Competition Law
↓
Data
↓
Network effects
↓
Platforms
↓
Algorithms
↓
Ecosystems

Intelligent Competition Law
↓
AI
↓
Autonomous optimisation
↓
Predictive systems
↓
Dynamic network effects
↓
Continuous market evolution
↓
Real-time competition regulation

This represents a fundamental conceptual shift.

31. Conclusion

Intelligent market evolution creates a dynamic dimension of antitrust law. Competition authorities can no longer examine only existing market shares and current prices. They increasingly need to understand how algorithms, data, network effects, interoperability, AI, and platform ecosystems alter competitive conditions over time.

The principal competition-law concern is not technological intelligence itself. Intelligent technologies can produce substantial efficiencies and innovation. The concern arises when technological advantages are transformed into mechanisms for exclusion, foreclosure, coordination, discrimination, or durable market power.

The central principle can therefore be stated as:

Competition law in intelligent markets must protect the competitive process not only against existing dominance, but also against technological mechanisms that can convert temporary advantages into self-reinforcing and durable market power.

The Google, Microsoft, Booking.com, Amazon, Facebook/Meta and Alibaba matters collectively demonstrate the movement from conventional market-power analysis toward algorithmic, data-driven, ecosystem-based and increasingly forward-looking competition analysis. Recent EU and Chinese enforcement developments show that this transition is continuing.