Competition Law and Antitrust Implications of Ecosystem Behavioural Intelligence

1. Introduction

Ecosystem Behavioural Intelligence (EBI) refers to the ability of a digital or commercial ecosystem to collect, aggregate, analyze, predict, and act upon behavioural information generated by users, businesses, suppliers, competitors, developers, advertisers, and other participants within the ecosystem.

It can involve:

consumer purchasing behaviour;

search behaviour;

browsing patterns;

supplier activity;

developer activity;

transaction histories;

pricing responses;

switching behaviour;

engagement patterns;

advertising interactions;

inventory information;

competitor performance;

platform usage;

algorithmically generated predictions.

EBI becomes particularly significant when a platform simultaneously observes an ecosystem and competes within that ecosystem.

For example, a dominant marketplace may observe millions of transactions and then use those insights to identify successful products, predict consumer demand, determine prices, rank sellers, develop competing products, and optimize its own logistics.

The central antitrust question is therefore:

When does legitimate intelligence derived from ecosystem participation become a mechanism for acquiring, maintaining, or exploiting market power?

Competition law does not generally prohibit the collection or analysis of commercial information by itself. The legal concern arises where behavioural intelligence is used in ways that produce exclusionary, discriminatory, collusive, exploitative, or anticompetitive effects.

2. Components of Ecosystem Behavioural Intelligence

EBI can be divided into several categories.

A. Consumer intelligence

Platforms may analyze:

searches;

purchases;

clicks;

abandoned transactions;

preferences;

willingness to pay;

switching patterns.

B. Supplier intelligence

A platform may know:

supplier prices;

inventory;

margins;

sales volumes;

product launches;

promotional strategies.

C. Competitor intelligence

An ecosystem operator can potentially observe:

rival product performance;

customer acquisition;

pricing;

availability;

advertising effectiveness.

D. Developer intelligence

App stores and software ecosystems can observe:

application downloads;

retention;

functionality;

revenue;

user engagement.

E. Algorithmic intelligence

AI can transform raw data into:

predictions;

rankings;

recommendations;

demand forecasts;

pricing suggestions;

churn predictions;

competitive assessments.

This transformation is important because the competitive advantage may arise not from possession of data alone but from the ability to convert ecosystem data into actionable intelligence.

3. Why EBI Creates Competition Concerns

A vertically integrated platform can occupy two positions simultaneously:

Platform operator → observes ecosystem participants

and

Competitor → competes with those same participants.

This creates a potential information asymmetry.

For example:

A marketplace learns that a third-party seller's product is rapidly gaining popularity. The marketplace then uses the seller's transaction data to launch a competing product and gives that product preferential placement.

Several competition theories could potentially arise:

leveraging;

self-preferencing;

discriminatory treatment;

misuse of commercially sensitive information;

exclusion;

data advantage;

tying;

margin squeeze;

vertical foreclosure.

The precise legal outcome would depend upon market power, conduct, evidence and effects.

4. Relevant Market

The relevant market in an EBI dispute could be:

online marketplace services;

digital advertising;

app distribution;

search services;

cloud services;

data analytics;

enterprise software;

payment services;

logistics;

digital identity;

social networking.

However, EBI can make traditional market definition difficult.

A platform may provide a service at zero monetary price while obtaining behavioural information in return.

Consequently, authorities may examine competition in terms of:

quality;

privacy;

innovation;

data access;

functionality;

switching costs;

network effects;

ecosystem participation.

5. Data Accumulation and Market Power

A platform can potentially develop a behavioural-data feedback loop:

More users → more behavioural data → better predictions → better service → more users → more data.

This can create:

economies of scope;

network effects;

learning effects;

entry barriers;

increasing returns to data.

However, possession of large quantities of data does not automatically establish dominance.

The relevant questions include:

Is the data commercially significant?

Is it difficult for competitors to obtain?

Can alternative data sources substitute for it?

Does the data materially improve the platform's service?

Does the platform possess durable market power?

Is the data advantage being used to exclude competitors?

6. United Brands v Commission

United Brands v Commission, Case 27/76

United Brands remains an important authority concerning the assessment of dominance.

The case demonstrates that market power must be evaluated through a combination of factors rather than through market share alone.

Application to EBI

A behavioural-intelligence platform may possess market power because of:

network effects;

user scale;

data advantages;

switching costs;

ecosystem integration;

access to unique behavioural information.

Thus, a competition authority could examine the overall competitive structure, rather than merely asking how much data a company possesses.

7. Google Shopping

The Google Shopping litigation is particularly relevant to ecosystem intelligence.

The underlying competition concern involved Google's treatment of its own comparison-shopping service within its general search ecosystem.

EBI relevance

A dominant ecosystem may possess extensive information concerning:

consumer preferences;

search queries;

click-through rates;

purchasing behaviour;

competing services.

If the ecosystem operator uses its control over the platform to systematically advantage its own downstream service, authorities may investigate whether this amounts to exclusionary conduct.

The broader lesson is that control over an information-rich platform can have competitive consequences when the operator also participates in downstream markets.

8. Amazon Marketplace Competition Issues

Amazon-related European competition proceedings are particularly relevant to ecosystem behavioural intelligence.

The competition concerns examined the use of marketplace data and the relationship between Amazon's marketplace operator role and its own retail activities.

EBI relevance

A marketplace can potentially observe:

seller sales;

product demand;

prices;

inventory;

customer behaviour;

product performance.

The competitive issue is whether such information is used in a way that disadvantages independent sellers.

The critical distinction is between:

legitimate platform analytics

and

use of competitively sensitive ecosystem information to obtain an exclusionary advantage.

9. Microsoft Corp. v Commission

Microsoft Corp. v Commission, Case T-201/04

Microsoft is important for understanding how technological control and interoperability can affect competition.

EBI relevance

A platform that controls a technological ecosystem can potentially observe and control:

application behaviour;

interoperability;

technical information;

developer activity.

If behavioural intelligence is combined with technical restrictions, the resulting conduct could potentially reinforce exclusionary effects.

For example, a platform might use knowledge concerning competing applications to modify APIs or technical requirements in ways that disadvantage them.

The legality would depend on the evidence and applicable legal test.

10. IMS Health v Commission

IMS Health GmbH & Co. OHG v NDC Health GmbH, Case C-418/01

IMS Health concerned the relationship between intellectual-property rights and access to commercially important structures.

EBI relevance

Behavioural intelligence may be embedded in:

databases;

analytical models;

proprietary classifications;

prediction systems;

software architecture.

Competition law does not normally require firms to surrender proprietary information merely because competitors want it.

The IMS Health framework illustrates the exceptional nature of compulsory access to protected assets.

11. Bronner v Mediaprint

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

Bronner established a demanding framework concerning refusal to provide access to infrastructure.

EBI relevance

Suppose a dominant platform possesses a unique behavioural-intelligence infrastructure that downstream businesses cannot realistically reproduce.

A refusal to provide access could raise questions concerning:

indispensability;

elimination of effective competition;

feasibility of duplication;

objective justification.

But mere usefulness of data does not automatically make it an essential facility.

12. Commercial Solvents

Commercial Solvents Corp. v Commission, Joined Cases 6/73 and 7/73

Commercial Solvents illustrates how a dominant undertaking controlling an important input can potentially abuse its position by restricting downstream competition.

EBI application

Imagine a dominant ecosystem controls a unique behavioural dataset needed by downstream competitors.

If the undertaking:

supplies information to affiliated businesses;

denies equivalent access to competitors; or

restructures access to foreclose downstream rivals,

competition authorities could investigate whether the conduct constitutes abusive exclusion.

13. Self-Preferencing

One of the most important EBI issues is self-preferencing.

A platform may use behavioural intelligence to improve its own downstream products.

For example:

Platform observes ecosystem → identifies profitable category → launches own product → uses platform data → ranks own product prominently.

Potential advantages include:

better product development;

better pricing;

improved advertising;

superior forecasting;

preferential ranking.

Self-preferencing is not inherently unlawful. The competition analysis depends on:

dominance;

conduct;

foreclosure;

effects;

objective justification;

applicable legislation.

14. Information Asymmetry

EBI can create a significant asymmetry between the ecosystem operator and ecosystem participants.

Consider:

Seller: sees its own sales.

Platform: sees the sales of thousands of sellers.

The platform may therefore know:

which products are becoming successful;

which prices attract consumers;

which sellers are losing customers;

which products have high margins;

which categories are expanding.

This information advantage may be competitively significant.

The key issue is whether the platform uses the asymmetry in a manner that undermines competition.

15. Use of Competitively Sensitive Information

Competition law has long been concerned with the exchange and use of competitively sensitive information.

Sensitive information may include:

future prices;

production quantities;

capacity;

strategic plans;

costs;

customer information;

sales forecasts.

In an ecosystem, the platform itself may acquire this information through ordinary operations.

The competition issue can arise when the platform uses that information to compete against the businesses from which it obtained it.

16. Algorithmic Behavioural Intelligence

Modern EBI systems may employ machine learning.

The system can predict:

customer switching;

demand;

willingness to pay;

competitor entry;

seller failure;

optimal prices.

This creates new competition questions.

Example

A dominant platform's algorithm determines that Seller A will raise prices tomorrow.

The platform immediately:

adjusts its own price;

increases its own inventory;

changes search ranking;

targets Seller A's customers.

Such conduct is not automatically unlawful.

However, where the information is obtained through a dominant platform and systematically used to disadvantage competitors, it may become relevant to an abuse-of-dominance analysis.

17. Algorithmic Collusion

EBI can also create risks among competitors.

Suppose competing firms use the same behavioural-intelligence provider.

The system receives:

competitor prices;

demand data;

inventory;

market conditions.

The algorithms may then optimize prices.

This can potentially facilitate:

coordination;

price alignment;

output restrictions;

market allocation.

Eturas UAB v Lithuanian Competition Authority

Case C-74/14

The CJEU examined an online platform through which a technical system could facilitate coordinated pricing behaviour.

EBI relevance

The case demonstrates why competition authorities may examine the technological environment through which commercial decisions are made.

The existence of common software, however, does not by itself establish a cartel.

Evidence concerning communication, awareness, participation and coordination remains important.

18. Pricing Personalization

EBI can enable highly individualized pricing.

A platform may predict each consumer's:

willingness to pay;

urgency;

likelihood of switching;

purchasing history.

The platform can then personalize offers.

Personalized pricing can produce efficiencies, but in a dominant ecosystem it can raise questions concerning:

exclusionary pricing;

discrimination;

exploitation;

transparency;

competitive foreclosure.

Competition law analysis should distinguish legitimate price differentiation from conduct designed to harm competitors or exploit market power.

19. Loyalty and Switching Intelligence

A platform can use behavioural data to identify customers who are likely to switch.

It could then provide those users with:

discounts;

targeted promotions;

exclusive offers;

additional services.

This can strengthen customer retention.

Where a dominant undertaking uses targeted incentives combined with exclusivity or contractual restrictions, authorities may examine whether the strategy forecloses competing suppliers.

20. Intel Corp. v Commission

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

Intel is a major authority concerning conditional rebates and potential exclusionary effects.

EBI application

Behavioural intelligence could allow a dominant platform to identify:

customers most likely to defect;

customers most important to rivals;

strategically important suppliers.

The platform could then target those users with individualized incentives.

The relevant competition question would be whether the pricing mechanism has exclusionary effects under the applicable legal framework.

21. Tying and Bundling

EBI can facilitate sophisticated bundling.

A platform may know that a customer heavily depends on:

cloud services;

advertising;

payment services;

analytics.

It could then bundle services strategically.

For example:

A dominant platform provides behavioural analytics only to customers purchasing its cloud infrastructure.

Potential issues include:

tying;

leveraging;

foreclosure;

increased switching costs.

22. Exclusive Dealing

Behavioural intelligence can make exclusive-dealing strategies more sophisticated.

Instead of offering the same incentive to every customer, the platform can identify:

which customer matters most to rivals;

which supplier is most contestable;

which developer is strategically important.

It can then offer targeted exclusivity incentives.

This could potentially increase foreclosure effects because resources are directed toward the competitors' most important trading partners.

23. Network Effects and Behavioural Data

EBI can reinforce network effects.

The cycle may be:

More users → more behavioural information → better algorithms → better service → greater user attraction → more users.

This can produce a significant competitive advantage.

A rival entering the market may have:

fewer users;

less behavioural data;

weaker predictive models;

fewer developers;

lower-quality recommendations.

This can create a self-reinforcing ecosystem.

However, authorities must distinguish genuine competition based on superior products from exclusionary conduct.

24. Data Portability

Data portability can reduce switching costs.

If customers can transfer:

transaction history;

preferences;

account information;

behavioural records;

digital profiles,

then competing platforms may find it easier to attract them.

Competition concerns may arise where a dominant undertaking intentionally:

makes exports difficult;

uses incompatible formats;

imposes excessive migration costs;

restricts APIs.

25. Interoperability

EBI becomes particularly powerful when combined with ecosystem interoperability.

A platform may control:

identity;

payments;

advertising;

cloud;

search;

application distribution.

If it prevents competitors from accessing important ecosystem functions, it can potentially reinforce market power.

The Microsoft jurisprudence is relevant because interoperability can determine whether competing products can effectively participate in an ecosystem.

26. Privacy as a Dimension of Competition

Behavioural intelligence creates an important connection between competition and privacy.

A platform might compete by offering:

stronger privacy;

less tracking;

reduced data collection.

Alternatively, a dominant platform might impose conditions requiring extensive data collection.

Competition authorities can consider privacy as a non-price competitive parameter in appropriate circumstances.

However, competition law should not automatically convert every privacy violation into an antitrust violation.

27. Ecosystem Expansion

EBI allows a platform to identify adjacent markets.

For example:

Search data → advertising → payments → shopping → logistics → financial services.

The platform can observe consumer behaviour across these markets.

This may provide an advantage when entering adjacent markets.

Potential competition concerns include:

leveraging;

tying;

cross-subsidization;

self-preferencing;

data combination;

exclusion of rivals.

28. Conglomerate Effects

A company controlling several complementary markets may use EBI across them.

For example:

Operating system + app store + payments + advertising + cloud.

Behavioural intelligence from one market can improve performance in another.

This creates possible ecosystem-wide economies of scope.

The competitive assessment should consider whether the resulting advantages arise from legitimate innovation or whether the platform uses dominance in one market to restrict competition elsewhere.

29. Merger Control

EBI can significantly affect merger analysis.

Consider a hypothetical acquisition:

A dominant marketplace acquires a behavioural analytics startup.

The startup may have low revenue but possess:

unique behavioural datasets;

advanced prediction models;

innovative AI;

valuable algorithms.

Traditional turnover thresholds may fail to capture the competitive importance of such an acquisition.

Authorities may therefore examine:

potential competition;

innovation;

data advantages;

future market development;

ecosystem expansion.

30. Killer Acquisitions

A dominant platform may acquire an emerging competitor primarily because its behavioural-intelligence technology could eventually challenge the incumbent.

Relevant factors may include:

user growth;

technological capabilities;

proprietary algorithms;

data assets;

innovation pipeline.

The competitive concern is potential foreclosure of future competition rather than simply current market share.

31. Standardization of Behavioural Data

Ecosystems may establish common data standards.

Standardization can improve:

interoperability;

portability;

innovation;

competition.

But dominant firms may attempt to control standards to:

exclude rivals;

make proprietary technology mandatory;

increase switching costs.

The competition assessment may therefore examine the standard-setting process itself.

32. Digital Advertising

Advertising is one of the markets most heavily influenced by behavioural intelligence.

A platform may know:

what users search;

what products they view;

what advertisements they click;

what purchases they make.

This can create a significant competitive advantage in advertising markets.

Potential concerns include:

combining datasets;

discriminatory access;

self-preferencing;

tying advertising services to other platform services;

restricting rival ad technologies.

33. Search and Ranking

A platform possessing ecosystem behavioural intelligence can optimize ranking algorithms.

Potential issues arise if the platform:

deliberately demotes competitors;

favors its own products;

manipulates rankings;

conditions visibility on purchasing additional services.

Ranking itself is not necessarily anticompetitive.

The relevant question is whether ranking practices by a dominant undertaking materially foreclose competing services.

34. App Stores

App stores provide an especially clear EBI environment.

The platform may observe:

app downloads;

purchases;

engagement;

revenue;

retention;

user reviews.

If the platform operates competing applications, it has potentially valuable intelligence about competitors.

Possible theories include:

self-preferencing;

discriminatory access;

tying;

commissions;

exclusionary contractual terms;

use of developer information.

35. Cloud Ecosystems

Cloud providers can observe substantial information about customers' technological activities.

A vertically integrated cloud provider could potentially use ecosystem intelligence to enter adjacent markets.

Potential competition questions include:

whether customers can migrate easily;

whether data is portable;

whether APIs are open;

whether competing software receives equivalent access;

whether cloud services are bundled with proprietary applications.

36. Competition Between Ecosystems

EBI can also create competition between entire ecosystems, rather than individual products.

Examples include:

mobile ecosystems;

cloud ecosystems;

e-commerce ecosystems;

enterprise software ecosystems;

advertising ecosystems.

The relevant competitive unit may therefore be an interconnected ecosystem.

The analysis should nevertheless identify the specific markets in which competitive harm allegedly occurs.

37. Objective Justifications

A platform may legitimately use ecosystem intelligence for:

fraud prevention;

cybersecurity;

product improvement;

demand forecasting;

quality control;

safety;

personalization;

infrastructure optimization.

Accordingly, the mere fact that a platform uses information generated by ecosystem participants should not automatically be treated as anticompetitive.

The crucial questions are:

What information is used?

How was it obtained?

For what purpose?

Does the platform possess dominance?

Does the practice foreclose rivals?

Are there legitimate efficiencies?

Are less restrictive alternatives available?

38. Possible Competition-Law Remedies

Where unlawful conduct is established, potential remedies may include:

Behavioural remedies

data-access obligations;

non-discrimination;

interoperability;

API access;

restrictions on use of competitively sensitive information;

transparency requirements.

Structural remedies

In exceptional cases:

divestiture;

separation of business units.

Merger remedies

licensing;

data-access commitments;

interoperability commitments;

firewalls;

divestiture of overlapping assets.

39. Important Case-Law Matrix

40. Ten Principal Antitrust Risks

The principal competition-law risks associated with ecosystem behavioural intelligence can therefore be summarized as:

1. Data foreclosure

Competitors cannot obtain comparable information.

2. Information exploitation

A platform uses participants' information to compete against them.

3. Self-preferencing

The platform uses behavioural intelligence to favor its own services.

4. Algorithmic discrimination

Different ecosystem participants receive strategically different treatment.

5. Exclusive dealing

Behavioural intelligence enables targeted exclusionary contracts.

6. Tying

The platform uses information advantages to sell complementary services.

7. Interoperability restrictions

Technical barriers prevent rivals from competing effectively.

8. Algorithmic coordination

Common intelligence systems facilitate coordination among competitors.

9. Data-driven acquisitions

A dominant firm acquires emerging competitive threats.

10. Ecosystem leveraging

Market power in one ecosystem layer is extended into adjacent markets.

41. Overall Legal Assessment

Ecosystem Behavioural Intelligence represents a significant development in digital competition because it transforms observation itself into a competitive resource.

The critical economic feature is the combination of:

Data + scale + algorithms + network effects + ecosystem control.

A platform that merely uses data to improve its services may be engaging in legitimate competition.

The competition-law problem becomes more significant where a dominant undertaking uses its informational position to:

exclude competitors;

discriminate against ecosystem participants;

restrict interoperability;

exploit commercially sensitive information;

impose loyalty or exclusivity;

favor its own downstream products;

facilitate coordination.

The cases of United Brands, Commercial Solvents, Bronner, IMS Health, Microsoft, Intel, Google Shopping, Eturas and Slovak Telekom provide a doctrinal foundation for addressing these issues.

Conclusion

Ecosystem Behavioural Intelligence can become both a source of competitive efficiency and a source of market power. Its importance lies not simply in the quantity of data collected but in the ability of an ecosystem operator to transform behavioural information into strategic advantages unavailable to independent competitors.

The most important future antitrust question is therefore likely to be whether control over ecosystem-generated intelligence becomes a durable competitive bottleneck.

Where EBI produces better products, lower costs, improved forecasting, innovation and consumer benefits, it may strengthen competition. Where a dominant undertaking uses the same intelligence to foreclose rivals, discriminate against dependent businesses, manipulate ecosystem access, or extend dominance into adjacent markets, established antitrust doctrines concerning abuse of dominance, refusal to deal, tying, exclusive dealing, self-preferencing, interoperability, information exchange and merger control become particularly relevant.