Competition Law and Competition Concerns in Digital Guidance Infrastructures

1. Introduction

Digital guidance infrastructures are digital systems that influence how users, businesses, or institutions find, select, rank, purchase, access, or interact with products and services. They include:

• search engines and vertical-search systems;
• recommendation and ranking engines;
• navigation and route-planning platforms;
• app-store discovery systems;
• digital assistants and AI agents;
• marketplace “Buy Box” and product-selection systems;
• travel, hotel, food-delivery and mobility recommendation platforms;
• financial or insurance comparison systems;
• algorithmic decision-support and allocation systems.

Competition concerns arise because the operator of the guidance infrastructure may not merely provide a neutral technical service. Where it possesses substantial market power, control over ranking, recommendations, visibility, data, defaults and access can become a means of influencing competition in adjacent markets.

This is particularly important in digital markets because users often do not inspect every available alternative. A platform's first recommendation, default result, ranking position or AI-generated answer may determine which competing products receive visibility.

China's Platform Economy Anti-Monopoly Guidelines expressly recognise concerns involving algorithms, data, platform rules, exclusive dealing, search downgrading and other forms of digital-platform conduct.

2. Meaning of Digital Guidance Infrastructure

A digital guidance infrastructure can be understood as having five interconnected layers:

A. Data layer

The system collects:

• search histories;
• purchase histories;
• location information;
• clicks;
• ratings;
• reviews;
• user preferences;
• seller performance;
• transaction data.

B. Algorithmic layer

Algorithms process the information to determine:

• rankings;
• recommendations;
• search results;
• suggested products;
• advertisements;
• route choices;
• default services.

C. Interface layer

The results are presented through:

• search-result pages;
• recommendation panels;
• default applications;
• AI responses;
• product cards;
• Buy Boxes;
• notifications.

D. Commercial layer

The platform may monetise guidance through:

• advertising;
• commissions;
• transaction fees;
• preferred placement;
• sponsored rankings;
• subscriptions;
• payments;
• cross-selling.

E. Ecosystem layer

The guidance system can direct users toward the platform's own:

• marketplace;
• payment service;
• logistics service;
• advertising service;
• cloud service;
• AI assistant;
• travel service;
• financial product.

The competition problem becomes particularly significant where one undertaking controls several of these layers simultaneously.

3. Principal Competition-Law Issues

3.1 Self-preferencing

The platform may place its own product or service above competing products.

Examples include:

platform-owned hotel service above rival hotel-comparison services;

platform-owned shopping results above competing comparison services;

platform-owned payment service receiving better technical integration;

platform-owned AI assistant receiving privileged access to device functions.

Self-preferencing becomes a competition concern when a dominant undertaking uses control over an essential discovery or guidance channel to disadvantage competitors.

The EU's Google Shopping decision is the leading illustration.

4. Relevant Legal Framework

European Union

Relevant provisions include:

• Article 102 TFEU;
• EU merger-control rules;
• Digital Markets Act;
• principles concerning exclusionary abuse;
• interoperability and access obligations under the DMA.

The DMA is particularly important because it addresses certain digital-gatekeeper practices directly rather than requiring every conduct to be established through traditional Article 102 analysis.

In July 2026, the European Commission fined Google €460 million concerning self-preferencing in Google Search and €430 million concerning Google Play steering. The Commission stated that Google had given preferential treatment to its own services, including shopping, hotels, transport and sports results.

China

Relevant legislation includes:

• Anti-Monopoly Law of the PRC;
• Anti-Monopoly Guidelines for the Platform Economy;
• provisions concerning abuse of dominance;
• refusal to deal;
• exclusive dealing;
• discriminatory treatment;
• tying;
• leveraging;
• algorithmic and data-related conduct;
• merger control.

China's Platform Economy Guidelines specifically recognise conduct such as platform-imposed “choose one from two” arrangements, search downgrading, traffic restrictions, technical barriers and algorithmic measures.

India

The Competition Act, 2002 is particularly relevant through:

• Section 3 — anti-competitive agreements;
• Section 4 — abuse of dominant position;
• Section 19 — investigation;
• Section 26 — investigation procedure;
• Section 27 — remedies.

The Google Android/Play Store decisions demonstrate how control over an important digital ecosystem can affect adjacent markets.

5. Major Competition Concerns

5.1 Ranking discrimination

A platform can manipulate rankings so that:

• its own products appear first;
• affiliated companies receive greater visibility;
• rivals are pushed down;
• independent sellers receive less traffic;
• certain business models receive preferential exposure.

This is particularly powerful where most users rarely move beyond the first few results.

The competition issue is therefore not simply whether ranking is technically different, but whether the ranking mechanism is being used to distort competitive conditions.

6. Important Case Laws

Case 1 — Google Search (Shopping) — European Union

Google Search (Shopping), European Commission

This is one of the most important cases for digital guidance infrastructure.

The European Commission found that Google had given its own comparison-shopping service more favourable positioning and display in general search results than competing comparison-shopping services.

The central competition concern was the interaction between:

1. Google's dominance in general search;
2. Google's control over ranking;
3. Google's own specialised shopping service;
4. visibility of competing services.

Google's search algorithms could demote competing comparison-shopping services, while Google's own service received preferential treatment.

The case demonstrates the concept of algorithmic self-preferencing.

The UK's competition authorities have also identified Google Shopping as a seminal example of how ranking algorithms can be used to favour a platform's own products.

Principle

A dominant digital intermediary may create competition concerns when it controls a critical discovery channel and uses that channel to advantage its own downstream service.

7. Case 2 — Google Android — European Union

Google Android, European Commission

The Android case concerned Google's conduct relating to the Android mobile ecosystem.

The Commission examined arrangements involving:

• Google Search;
• Google Chrome;
• Google Play;
• Android device manufacturers;
• mobile application distribution.

The broader significance for digital guidance infrastructure is that defaults and ecosystem integration can influence user choice.

A user does not necessarily make a fully independent choice between every search engine, browser or application. Default placement and ecosystem architecture can substantially affect discoverability.

Competition significance

The case demonstrates that competition law can examine the interaction between:

operating system → default → search → user data → advertising → downstream markets.

8. Case 3 — Google Android / UPI — Competition Commission of India

Umar Javeed & Others v. Google LLC & Another

CCI Case No. 39/2018

The CCI examined Google's Android ecosystem and related conduct.

The CCI imposed a penalty of approximately ₹1,337.76 crore in October 2022 for anti-competitive practices relating to Android mobile devices.

The decision is relevant to digital guidance infrastructure because control over Android and associated services could affect:

• search;
• browser choice;
• application distribution;
• payment applications;
• competing digital services.

The CCI also examined differential treatment of Google's own UPI application compared with competing UPI applications.

The Commission considered that Google's position in Android app distribution gave it a responsibility not to provide its own application with an unjustified competitive advantage.

Principle

Control over a digital ecosystem can give the platform the ability to guide users toward affiliated services, making discriminatory technical integration a potential competition concern.

9. Case 4 — Google Play Store — India

XYZ v. Alphabet Inc. & Others / Google Play Store cases

The CCI's 2022 Google Play Store decision concerned Google's policies affecting app developers and payment-processing services.

The CCI found concerns involving:

• mandatory use of Google's billing system;
• anti-steering restrictions;
• access to alternative payment systems;
• leveraging of Google's position;
• discrimination affecting competing services.

The CCI imposed a penalty of approximately ₹936.44 crore.

The decision is significant for digital guidance infrastructures because an app marketplace does not simply host applications. It determines:

• which applications users can discover;
• how applications are distributed;
• how payments are made;
• how developers communicate with users;
• which commercial channels are available.

Principle

Control over digital discovery plus transaction infrastructure can enable a platform to extend power from one market into adjacent markets.

10. Case 5 — Amazon Marketplace — UK CMA

Competition and Markets Authority — Investigation into Amazon's Marketplace

The UK's CMA investigated Amazon's marketplace concerning:

• third-party seller data;
• selection of the Amazon Buy Box;
• delivery-rate negotiations for Prime orders.

The CMA accepted commitments from Amazon addressing these concerns.

The Buy Box is particularly relevant to digital guidance infrastructure because it determines which seller's offer is prominently displayed when a consumer reaches a product page.

Consequently, the ranking mechanism can have substantial commercial significance.

Competition significance

The case illustrates the possibility that:

algorithmic selection → visibility → consumer choice → sales allocation

can become a competition issue.

The FTC's record similarly describes the importance of Amazon's Buy Box, noting that Amazon uses criteria including price, delivery speed, Prime eligibility and seller performance.

11. Case 6 — Amazon Marketplace — United States

FTC & State Attorneys General v. Amazon

The FTC and several states sued Amazon alleging that Amazon maintained monopoly power through various interconnected practices.

The allegations include conduct concerning:

• marketplace sellers;
• pricing;
• advertising;
• product visibility;
• seller relationships;
• Amazon's marketplace ecosystem.

The FTC alleges that Amazon's practices can restrict rivals and prevent competing platforms from achieving sufficient scale.

The case is particularly relevant to digital guidance because Amazon's marketplace simultaneously performs several functions:

discovery + ranking + advertising + transaction + fulfilment.

Where those functions are vertically integrated, the platform may have incentives and opportunities to favour its own ecosystem.

Important qualification

These are allegations in ongoing litigation, not final judicial findings on every alleged practice. The FTC case remained pending as of August 2026.

12. Case 7 — Meituan — China

Meituan “Choose One from Two” Case

China's SAMR investigated Meituan's conduct in the online food-delivery platform market.

SAMR found that Meituan used mechanisms including:

• differentiated fees;
• delayed onboarding;
• exclusive cooperation;
• deposits;
• data;
• algorithms;
• punitive measures.

The investigation concluded that these mechanisms supported an exclusive-dealing arrangement that restricted competition. SAMR imposed a fine of approximately RMB 3.442 billion and required corrective measures concerning platform rules and algorithms.

Relevance

The case demonstrates that algorithmic visibility and access can be connected to exclusionary arrangements.

A platform need not explicitly tell a merchant:

“You cannot use a competing platform.”

It can potentially use:

• search downgrading;
• traffic restrictions;
• delayed listing;
• discriminatory algorithms;
• differential fees

to produce similar competitive effects.

13. Case 8 — Alibaba — China

Alibaba Group — Online Retail Platform Case

SAMR found Alibaba to have abused its dominant position in China's online retail platform services market through an exclusive-dealing practice commonly described as “choose one from two.”

Alibaba was fined approximately RMB 18.228 billion, equivalent to 4% of its 2019 domestic sales, and was required to undertake corrective measures.

Relevance to digital guidance infrastructure

Alibaba's platform ecosystem contained multiple mechanisms capable of influencing merchant visibility and commercial opportunities.

The case demonstrates that platform rules, incentives and penalties can collectively influence where sellers participate and consequently what products consumers can discover.

Principle

Digital guidance cannot be examined independently from the broader platform ecosystem when the guidance system determines traffic allocation and commercial access.

14. Case 9 — Tencent Music — China

Tencent / China Music Group

SAMR investigated Tencent's acquisition of China Music Group.

The relevant market was identified as China's online music-streaming platform market. SAMR found that the transaction gave the combined entity control over a very large share of exclusive music-copyright resources.

SAMR required Tencent to:

• terminate exclusive copyright arrangements;
• stop certain high-prepayment practices;
• avoid obtaining preferential conditions from copyright owners without justification.

Relevance

Although this was primarily a merger-control case rather than a pure ranking case, it is relevant to digital guidance infrastructure because access to content is a fundamental input into:

• recommendation engines;
• search;
• playlists;
• personalised discovery;
• music rankings.

If a platform controls a critical content input, competitors may find it difficult to offer comparable guidance and recommendation services.

15. Case 10 — Google Search and DMA — European Union, 2026

The recent EU DMA enforcement against Google is particularly significant for the future of digital guidance infrastructures.

In July 2026, the European Commission concluded that Google had breached DMA obligations by giving preferential treatment to its own services in Google Search, including:

• shopping;
• hotels;
• transport;
• sports.

The Commission imposed a €460 million fine for the Search self-preferencing issue.

This is important because the DMA expressly addresses ranking neutrality.

The regulatory model therefore moves beyond asking only whether traditional Article 102 requirements have been established and imposes specific obligations on designated gatekeepers.

16. AI Assistants as New Digital Guidance Infrastructure

The next generation of competition concerns involves AI assistants.

An AI assistant may answer:

“Which hotel should I book?”

“Which laptop should I buy?”

“Which payment service should I use?”

“Which route should I take?”

“Which insurance policy is suitable?”

Unlike conventional search, the AI system may provide one synthesised recommendation rather than a list of ten results.

This creates several competition concerns.

A. Answer monopolisation

If one AI assistant becomes the principal gateway to consumer decisions, visibility may shift from search-result competition to answer-level competition.

B. Self-preferencing

An AI provider might recommend:

• its own products;
• affiliated services;
• preferred advertisers;
• ecosystem partners.

C. Data advantage

The AI provider may possess:

• search data;
• transaction data;
• behavioural data;
• conversational data;
• location information.

D. Training-data advantage

Large datasets may improve recommendation quality and create barriers to entry.

E. Interoperability

Competitors may require access to:

• operating-system functions;
• APIs;
• search data;
• device capabilities;
• application actions.

The European Commission in 2026 issued binding DMA specification measures concerning Google's Android interoperability for competing AI services and access to Google Search data for third-party search engines.

17. Algorithmic Discrimination

Digital guidance infrastructure can discriminate between equivalent businesses.

Possible variables include:

The legal question is not whether ranking differentiation is inherently unlawful. Differentiation is often legitimate.

The important questions are:

1. Is the undertaking dominant?
2. What is the relevant market?
3. What is the ranking mechanism?
4. Is the criterion objectively justified?
5. Does the mechanism disadvantage competitors?
6. Does it foreclose efficient competitors?
7. Is there a consumer or technical justification?
8. Is the effect substantial enough to harm competition?

18. Data as a Competitive Advantage

Digital guidance systems become more powerful as they accumulate data.

For example:

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

This can produce a feedback loop.

A dominant platform can potentially use data from:

• search;
• advertising;
• transactions;
• sellers;
• payments;
• logistics;
• user interactions

to improve its recommendation system.

The competition concern is therefore not merely “data ownership.”

It is whether exclusive access to competitively important data creates durable market power or enables exclusionary conduct.

19. Network Effects

Digital guidance systems frequently exhibit network effects.

Direct network effects

More users can make the platform more valuable.

Indirect network effects

More sellers attract users, while more users attract sellers.

Data network effects

More interactions produce more data, which may improve recommendations.

Reputation effects

A platform with more historical information may generate more accurate rankings.

This can create a cycle:

Users → Data → Better Guidance → More Users → More Data.

A competitor may therefore face difficulty entering even if it has technically comparable software.

20. Tying and Bundling

A digital guidance infrastructure can be tied to another service.

Examples:

• search tied to browser;
• app distribution tied to payment;
• AI assistant tied to operating system;
• navigation tied to mapping;
• marketplace tied to logistics;
• hotel search tied to booking;
• shopping recommendation tied to advertising.

Competition law may become relevant where a dominant undertaking uses power in one market to extend its position into another.

21. Default Settings

Defaults are particularly important.

Consumers frequently accept:

• default search engines;
• default browsers;
• default navigation applications;
• default payment systems;
• default AI assistants.

Even where alternative products remain technically available, default positioning can significantly affect actual user behaviour.

Therefore, competition analysis should distinguish:

formal choice from effective choice.

A consumer may technically be free to select another provider while practical switching costs make that alternative much less visible.

22. Switching Costs and Lock-In

Digital guidance systems can create lock-in through:

• saved preferences;
• transaction histories;
• loyalty programmes;
• personalised recommendations;
• accumulated reviews;
• proprietary data;
• subscription benefits;
• ecosystem integration.

A user may remain with a platform not because alternatives are unavailable but because switching means losing accumulated digital value.

This can reduce competitive pressure.

23. Refusal to Provide Access

Competition issues may also arise where a dominant guidance infrastructure refuses competitors access to:

• APIs;
• search indexes;
• device functions;
• interoperability interfaces;
• marketplace data;
• technical standards;
• essential datasets.

The relevant question is whether the requested resource is genuinely indispensable and whether refusal can foreclose effective competition.

The EU's 2026 Google Android and Search-data measures illustrate the growing importance of interoperability and data access in digital competition regulation.

24. Algorithmic Collusion

Guidance infrastructure can also facilitate coordination.

For example, competing platforms may use algorithms that:

• monitor competitors;
• react automatically to price changes;
• synchronise recommendations;
• adjust rankings;
• respond to competitor inventory.

This creates a distinction between:

Explicit coordination

Humans agree to coordinate.

Algorithm-assisted coordination

Humans design systems capable of facilitating coordination.

Autonomous algorithmic coordination

Algorithms interact and potentially converge without conventional human communication.

Competition authorities therefore increasingly need to examine algorithm design, training objectives, data inputs and monitoring systems, not merely communications between executives.

25. Advertising and Sponsored Recommendations

A platform may combine:

organic recommendation + advertising + ranking.

This can create conflicts where commercial incentives affect supposedly neutral guidance.

For example:

A platform may have an incentive to recommend a product that generates greater advertising revenue rather than the product that would otherwise receive the highest organic ranking.

Competition concerns can arise from:

• discriminatory ad auctions;
• preferential placement;
• opaque sponsored ranking;
• exclusionary advertising requirements;
• tying advertising expenditure to visibility.

The distinction between paid placement and organic recommendation is therefore increasingly important.

26. Consumer Choice as a Competition Dimension

Traditional competition analysis often focuses on:

• price;
• output;
• quality.

Digital guidance infrastructure requires additional attention to:

• visibility;
• discoverability;
• ranking;
• interoperability;
• innovation;
• privacy;
• switching costs;
• data access;
• consumer autonomy.

A competitor may technically remain in the market while becoming commercially invisible.

Thus:

Market access without meaningful discoverability may not amount to effective competition.

27. Remedies

Competition authorities can use several remedies.

1. Non-discrimination

Require equal treatment of affiliated and independent services.

2. Ranking transparency

Require disclosure of material ranking parameters.

3. Choice screens

Give users meaningful opportunities to select alternative services.

4. Interoperability

Allow rival services access to necessary technical functions.

5. Data access

Provide competitors with appropriate access to competitively important data.

6. Anti-steering restrictions

Prevent platforms from stopping businesses from communicating alternative offers.

7. Structural separation

In exceptional cases, separate infrastructure from downstream commercial operations.

8. Algorithmic auditing

Independent review of ranking and recommendation systems.

9. Data-use restrictions

Prevent a platform from using third-party seller data to disadvantage those sellers.

10. Monitoring trustees

Continuous monitoring of compliance with behavioural remedies.

28. Key Doctrinal Tests

For an exam or legal analysis, the following framework can be applied:

Step 1 — Define the relevant market

Identify:

• product/service market;
• geographic market;
• platform-side markets;
• multi-sided market relationships.

Step 2 — Determine market power

Consider:

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

Step 3 — Identify the guidance function

Ask whether the platform controls:

• ranking;
• search;
• recommendations;
• defaults;
• AI answers;
• transaction allocation.

Step 4 — Identify the conduct

Possible conduct includes:

• self-preferencing;
• discrimination;
• tying;
• refusal to deal;
• exclusion;
• algorithmic manipulation;
• exclusive dealing;
• data exploitation.

Step 5 — Examine competitive effects

Determine whether the conduct:

• forecloses rivals;
• raises rivals' costs;
• restricts market access;
• reduces innovation;
• increases switching costs;
• reduces consumer choice.

Step 6 — Consider justification

Potential legitimate explanations include:

• quality;
• relevance;
• security;
• fraud prevention;
• technical efficiency;
• user experience;
• privacy;
• reduction of transaction costs.

Step 7 — Consider proportionality

The regulatory question should be whether the restriction is necessary and proportionate to the legitimate objective.

29. Comparative Case-Law Table

30. Emerging Competition Risks

The future of digital guidance infrastructure is moving from search results to machine-generated decisions.

The major emerging risks are:

1. AI self-preferencing
2. AI shopping recommendations
3. AI travel recommendations
4. AI financial-product recommendations
5. AI advertising allocation
6. algorithmic ranking discrimination
7. exclusive access to training data
8. AI-agent interoperability
9. platform-controlled defaults
10. autonomous algorithmic coordination
11. recommendation manipulation
12. vertical integration between guidance and transaction services

The shift is therefore:

Search → Ranking → Recommendation → Prediction → AI-generated decision.

The closer the digital system comes to making the consumer's decision itself, the more important competition rules concerning neutrality, interoperability, transparency and non-discrimination become.

31. Conclusion

Digital guidance infrastructures occupy a strategically important position in modern digital markets because they determine what users see, what they discover, which businesses receive traffic and which services become commercially viable.

The principal competition-law concerns are:

• self-preferencing;
• discriminatory ranking;
• algorithmic exclusion;
• manipulation of recommendations;
• tying and bundling;
• control of defaults;
• refusal of interoperability;
• data advantages;
• exclusive dealing;
• network effects;
• switching costs;
• algorithmic coordination.

The cases involving Google, Amazon, Alibaba and Meituan demonstrate that competition authorities increasingly examine not only traditional prices and market shares but also algorithms, data, ranking, visibility, ecosystem control and digital access. China's Meituan decision is especially illustrative of how algorithms and platform mechanisms can support exclusionary conduct, while the EU's recent Google DMA enforcement demonstrates the growing move toward explicit regulation of self-preferencing in digital guidance systems.

For legal analysis, the central principle is:

Control over a digital guidance channel can become a source of market power when competitors depend upon that channel to reach users.

Accordingly, competition law must examine not only who controls the market, but also who controls the pathway through which consumers discover and select competing products and services.