Ad-Tech Ecosystem Concentration And Attention Allocation Control .

Ad-Tech Ecosystem Concentration and Attention Allocation Control in Europe

1. Introduction

Ad-tech ecosystem concentration and attention allocation control concerns the accumulation of economic and technological power by a small number of digital platforms over the process through which users' attention is measured, ranked, allocated, monetised and sold to advertisers.

The modern ad-tech ecosystem can include:

search engines;

social-media platforms;

video platforms;

app stores;

operating systems;

demand-side platforms (DSPs);

supply-side platforms (SSPs);

ad exchanges;

publisher ad servers;

identity and tracking systems;

data-management systems;

recommendation engines;

real-time bidding systems;

measurement and attribution tools.

The central competition-law problem is not merely that a company has a large advertising business. The more difficult issue arises where the same ecosystem can potentially control multiple stages of the attention and advertising chain while also competing with businesses that depend upon that chain.

The European approach therefore examines market power, ecosystem effects, self-preferencing, data accumulation, foreclosure, discriminatory access, interoperability, ranking, advertising intermediation and conflicts of interest.

The most important legal provisions are:

Article 101 TFEU — anticompetitive agreements;

Article 102 TFEU — abuse of dominance;

Digital Markets Act (DMA);

national competition legislation;

national digital-market provisions such as Section 19a GWB in Germany;

GDPR where personal-data processing forms part of the competitive conduct;

consumer and platform-transparency legislation.

As of 2026, Alphabet's online advertising service is expressly designated as a DMA core platform service, and the Commission also adopted a September 2025 Article 102 decision concerning Google's ad-tech practices. (Digital Markets Act (DMA))

2. Meaning of Attention Allocation Control

Attention allocation

"Attention allocation" refers to the platform's ability to determine which content, advertisements, products or services receive visibility and which receive less visibility.

For example:

User → search/recommendation algorithm → ranking → screen position → user attention → advertising impression → advertiser payment

The platform may influence:

ranking;

recommendation;

search results;

advertising placement;

auction participation;

frequency of advertisements;

targeting;

personalised content;

notifications;

default settings;

visibility of competing services.

Thus, attention can function as an economically valuable input.

3. What Is an Ad-Tech Ecosystem?

A simplified ecosystem looks like this:

Advertiser

↓

Demand-Side Platform

↓

Ad Exchange / Auction

↓

Supply-Side Platform

↓

Publisher / App / Website

↓

User

At the same time, a large digital ecosystem may own several layers.

For example:

Operating System

→ Browser

→ Search Engine

→ Video/Social Platform

→ Advertising Exchange

→ Publisher Tools

→ Analytics

→ Advertising Data

This creates the possibility of vertical integration.

Vertical integration itself is not unlawful.

The competition question is:

Does control over one layer allow the undertaking to disadvantage rivals at another layer?

4. Why Ecosystem Concentration Matters

Traditional competition analysis often focuses on a defined product market.

Digital ecosystems make that approach more complicated because one company may simultaneously operate across several related markets.

The Google Android judgment expressly addressed the concepts of multi-sided platforms and ecosystems, demonstrating the importance of analysing interconnected digital markets rather than viewing each service in complete isolation. (Infocuria)

An ecosystem can create:

Network effects

More users → more data → better targeting → more advertisers → more revenue → more investment → more users.

Data advantages

More behavioural data may improve:

targeting;

recommendation;

measurement;

advertising prediction;

auction optimisation.

Switching costs

Advertisers and publishers may become dependent upon:

historical campaign data;

audience data;

technical integrations;

APIs;

measurement tools.

Economies of scope

The same infrastructure can support multiple services.

5. The "Attention Bottleneck"

A particularly important theory is the attention bottleneck.

Suppose a platform controls a large share of:

search queries;

social-media feeds;

video consumption;

mobile-device interfaces.

It may possess the ability to influence where users look and what they see.

That does not automatically establish an infringement.

However, competition concerns can arise where the platform uses that control to:

favour its own advertising services;

disadvantage rival ad-tech intermediaries;

restrict access to data;

manipulate ranking;

impose discriminatory technical conditions;

tie services together;

restrict interoperability;

prevent competitors from reaching users efficiently.

6. Article 102 TFEU

Article 102 prohibits abuse of a dominant position where conduct affects trade between EU Member States.

Relevant forms of conduct can include:

exclusionary agreements;

tying;

refusal of access;

discriminatory conditions;

self-preferencing;

exploitative conduct;

leveraging dominance from one market into another.

The crucial point is:

Dominance is not itself illegal. Abuse of that dominance is the prohibited conduct.

7. Digital Markets Act

The DMA adds a different regulatory model.

Rather than requiring the Commission always to prove traditional Article 102 dominance and effects, designated gatekeepers face specific obligations.

Alphabet's designated core-platform services include:

Google Search;

YouTube;

Google Play;

Google Shopping;

Android;

Chrome;

Google Maps;

Alphabet's online advertising service. (Digital Markets Act (DMA))

For advertising, the DMA contains important transparency requirements.

Advertisers can request information concerning:

prices;

fees;

remuneration;

calculation metrics.

Publishers receive corresponding transparency rights, and advertisers and publishers can obtain access to certain performance-measurement tools and relevant data for independent verification. (Digital Markets Act (DMA))

This is important because control over advertising information can itself create competitive dependence.

8. Self-Preferencing

One of the central theories is self-preferencing.

Example:

Platform operates an ad exchange + publisher ad server + competing ad marketplace.

If the platform systematically gives its own advertising service:

better access;

better ranking;

superior information;

faster processing;

preferential auction treatment;

while disadvantaging rivals, competition law may become relevant.

Self-preferencing became particularly important following the Google Shopping litigation.

9. Case Law 1 — Google Shopping

Google and Alphabet v Commission, C-48/22 P

Court: Court of Justice of the European Union
Judgment: 10 September 2024
ECLI: EU:C:2024:726

Facts

Google operated a dominant general search engine and also operated its own comparison-shopping service.

The Commission found that Google gave its own comparison-shopping results preferential positioning and display.

Judgment

The CJEU upheld the finding of abuse and the €2.4 billion fine.

The Court accepted that Google's conduct could constitute an abuse where its dominant search position was leveraged to favour its own specialised service and foreclose competition. (Infocuria)

Relevance to ad-tech

The case is highly relevant to attention allocation because it establishes that control over a high-traffic interface can be used to advantage another service.

The broader lesson is:

Visibility and ranking can constitute an important competitive parameter.

This is directly relevant where a platform controls both the interface through which attention is obtained and the monetisation system through which that attention is sold.

10. Case Law 2 — Google Android

Google and Alphabet v Commission, T-604/18

Court: General Court
Judgment: 14 September 2022
ECLI: EU:T:2022:541

Facts

The case concerned Google's Android ecosystem.

The Commission challenged several practices involving:

Google Search;

Chrome;

Play Store;

Android;

device manufacturers;

mobile network operators.

Judgment

The General Court largely upheld the Commission's findings that Google had imposed unlawful restrictions that helped consolidate the dominant position of Google Search.

The Court expressly dealt with:

multi-sided platforms;

ecosystems;

tying;

exclusivity payments;

anti-fragmentation obligations;

exclusionary effects. (Infocuria)

Relevance

The case demonstrates how control over an upstream digital layer—the operating system—can influence competition in downstream services.

For attention markets:

Control over the device/interface can influence access to users and therefore access to advertising attention.

11. Case Law 3 — Google AdSense for Search

Google and Alphabet v Commission, T-334/19

Court: General Court
Judgment: 18 September 2024
ECLI: EU:T:2024:634

This is one of the most directly relevant European ad-tech authorities.

Facts

The case concerned Google's position in the market for online search advertising intermediation.

Google's agreements with certain publishers contained contractual restrictions concerning competing search advertisements.

The Commission had imposed a fine of approximately €1.49 billion.

Judgment

The General Court annulled the Commission decision because it found errors in the Commission's assessment of the duration and market coverage of the contractual clauses and therefore concluded that the abuse had not been established as required.

The Court nevertheless provided important legal analysis concerning exclusive-supply obligations and exclusionary effects in online advertising intermediation. (Infocuria)

The Commission appealed, and C-826/24 P remains pending before the CJEU as of September 2026. The appeal was heard in July 2026. (Infocuria)

Importance

This case demonstrates an essential point:

Ad-tech concentration must be demonstrated through rigorous evidence of foreclosure and competitive effects; market importance alone is insufficient.

12. Case Law 4 — Meta Platforms v Bundeskartellamt

Meta Platforms and Others, C-252/21

Court: CJEU, Grand Chamber
Judgment: 4 July 2023
ECLI: EU:C:2023:537

Facts

Meta collected information from:

Facebook;

Instagram;

WhatsApp;

third-party websites;

third-party applications.

That information could be combined to create detailed profiles used in Meta's advertising-funded business model. (curia)

Judgment

The CJEU held that a competition authority can take account of possible GDPR infringements when assessing abuse of dominance, while respecting the institutional role of data-protection authorities.

Relevance to attention allocation

This case is fundamental because modern advertising power depends heavily on:

Data → profiling → prediction → targeting → attention → advertising revenue.

Therefore:

Data governance can become a competition issue when control over data strengthens a dominant platform's advertising position.

The German Bundeskartellamt subsequently closed the Facebook proceeding in October 2024 after measures were implemented giving users greater choices regarding data combination. (Federal Cartel Office)

13. Case Law 5 — Amazon Digital Ecosystem

Bundeskartellamt Amazon proceedings

Germany provides an important ecosystem-based approach under Section 19a GWB.

In July 2022, the Bundeskartellamt determined that Amazon had paramount significance for competition across markets.

The authority considered Amazon's:

marketplace;

retail business;

Prime ecosystem;

advertising;

payment;

logistics;

cloud services;

data resources.

The authority specifically identified Amazon's ability to act as both marketplace operator and competitor to sellers as an important structural feature. (Federal Cartel Office)

Attention relevance

Amazon is particularly important because visibility is economically valuable.

Amazon can influence:

search ranking;

Buy Box;

product visibility;

advertising eligibility.

In 2025, the Bundeskartellamt stated in preliminary findings that Amazon's price-control mechanisms could restrict the visibility of sellers' offers, including by affecting Buy Box and search-result display and potentially advertising eligibility. (Federal Cartel Office)

This demonstrates how price rules and attention allocation can become interconnected.

14. Case Law 6 — Amazon Seller Terms

Bundeskartellamt, B2-88/18

The German competition authority investigated Amazon's treatment of third-party sellers.

The investigation concerned:

account blocking;

payment withholding;

reviews;

liability clauses;

information rights;

contractual terms;

dependence upon Amazon's marketplace.

Amazon subsequently amended its terms and the proceedings were closed in 2019. (Federal Cartel Office)

Importance

The case demonstrates that digital-platform power can arise not merely through formal market share but through economic dependence.

For attention markets, a publisher or advertiser may technically have alternatives while practically remaining dependent on a dominant platform because access to its users is difficult to replace.

15. Case Law 7 — Apple App Tracking Transparency

Bundeskartellamt proceedings concerning Apple's ATTF

In February 2025, the Bundeskartellamt issued a preliminary legal assessment concerning Apple's App Tracking Transparency Framework (ATTF).

The authority's preliminary view was that Apple imposed stricter requirements on third-party applications than on itself concerning access to advertising-related data.

The authority identified concerns involving:

unequal treatment;

self-preferencing;

first-party versus third-party tracking;

consent interfaces;

advertising data;

app publishers;

advertisers;

ad-tech providers. (Federal Cartel Office)

Importance

This is a particularly strong illustration of attention allocation control through data access.

If a platform restricts competitors' ability to collect advertising data while preserving greater access for itself, it can potentially affect:

measurement → targeting → ad effectiveness → advertiser demand → publisher revenue → competitive position.

The assessment was preliminary, not a final finding of infringement.

16. Case Law 8 — Amazon's Price-Control and Visibility Proceedings

The Bundeskartellamt's 2025 Amazon assessment also provides a useful modern example.

The authority stated that Amazon's price-control mechanisms could cause certain offers to:

disappear from the Buy Box;

receive reduced search visibility;

become excluded from Amazon advertising.

The authority's preliminary assessment considered whether those practices could violate Section 19a GWB, Section 19 GWB and Article 102 TFEU. (Federal Cartel Office)

This illustrates an important emerging concept:

Platform governance can simultaneously regulate price, ranking, visibility and advertising access.

That combination can be more significant than any individual rule considered separately.

17. Concentration of the Ad-Tech Stack

The major competition concern is stack concentration.

A simplified concentrated ecosystem could look like:

User data

↓

Identity

↓

Audience profiling

↓

Advertiser demand

↓

Ad exchange

↓

Auction

↓

Publisher supply

↓

Measurement

If the same corporate group controls several of these layers, it may have:

information advantages;

conflict-of-interest incentives;

ability to discriminate;

ability to favour its own services;

ability to restrict rivals;

ability to determine access conditions.

This is particularly significant in real-time advertising.

18. Conflict of Interest

Suppose an undertaking simultaneously acts as:

advertiser intermediary;

ad exchange;

publisher ad server;

publisher;

auction operator;

measurement provider.

There may be a structural conflict.

The undertaking potentially possesses information concerning:

bids;

winning prices;

advertiser demand;

publisher inventory;

rival intermediaries;

user behaviour.

Competition law may therefore ask:

Can the vertically integrated undertaking use information or control over one layer to disadvantage rivals at another layer?

The Commission's 2025 Google ad-tech decision directly addressed this type of concern. The Commission found Google had favoured its own online display-advertising technology services and imposed a €2.95 billion fine, while ordering measures addressing the identified conflicts of interest. (European Commission)

19. Attention Allocation Through Ranking

Ranking can operate as a competitive gatekeeper.

Consider:

Position 1 — 40% of clicks

Position 2 — 20%

Position 3 — 10%

and so forth.

Even without formally excluding a rival, moving it from position 1 to position 10 may substantially reduce its ability to reach users.

Therefore, competition analysis may examine:

ranking algorithms;

search placement;

recommendation;

default positions;

sponsored placement;

organic placement;

preferential treatment;

traffic diversion.

Google Shopping is especially important here because the CJEU accepted that preferential display of Google's own service could constitute an abuse in the circumstances established by the case. (curia)

20. Attention Allocation Through Recommendations

Social media and video platforms can control attention through recommendation algorithms.

The system may determine:

which video appears next;

which post appears first;

which creator becomes visible;

which advertisement appears;

how frequently the user sees particular content.

This can create competition concerns where an undertaking:

favours its own commercial services;

disadvantages rival services;

uses data unavailable to rivals;

restricts interoperability;

manipulates access conditions.

However, algorithmic ranking by itself is not unlawful.

There must be an applicable competition-law or regulatory basis.

21. Data as a Competitive Input

Modern ad-tech competition can be represented as:

Data → Prediction → Targeting → Attention → Conversion → Revenue

Data can therefore function as an important competitive input.

Relevant categories include:

browsing data;

search data;

location data;

purchase data;

engagement data;

viewing history;

device information;

app activity;

cross-service activity.

The Meta judgment shows how data combination can be analysed simultaneously through competition law and data-protection law. (curia)

22. Network Effects

Ad-tech ecosystems may produce several network effects.

User side

More users → more behavioural data.

Advertiser side

More users → greater advertising reach.

Publisher side

More advertisers → potentially greater monetisation.

Data side

More interactions → better prediction.

Algorithm side

More data → better optimisation.

This can create a reinforcing cycle:

Users → Data → Better targeting → Advertisers → Revenue → Investment → More users

A competition authority may therefore examine whether network effects create durable entry barriers.

23. Switching Costs

Advertisers and publishers may incur significant switching costs.

Examples:

historical campaign data;

conversion tracking;

audience segments;

API integrations;

reporting systems;

billing systems;

attribution models;

technical infrastructure.

A platform need not prohibit switching expressly if the practical cost of leaving is sufficiently high to create dependence.

24. Multi-Homing

An important defence against ecosystem power is multi-homing.

An advertiser can potentially use:

Google Ads;

Meta Ads;

Amazon Ads;

independent DSPs;

publisher-direct advertising.

If advertisers can easily use multiple platforms, market power may be constrained.

But multi-homing may be limited by:

technical integration costs;

data fragmentation;

attribution difficulties;

audience duplication;

lack of interoperability;

exclusive arrangements.

25. Self-Preferencing vs Legitimate Product Design

Not every preferential result is illegal.

A platform may legitimately improve:

quality;

security;

relevance;

fraud prevention;

user experience.

The legal question is whether the conduct constitutes prohibited exclusionary behaviour or violates a specific DMA obligation.

The Google Shopping judgment demonstrates that the assessment requires attention to the competitive effects and the circumstances of the conduct, rather than treating every form of self-preferencing as automatically unlawful. (Infocuria)

26. Foreclosure

Foreclosure occurs where conduct makes it more difficult for competitors to compete effectively.

In ad-tech, foreclosure can occur through:

exclusive supply agreements;

tying;

discriminatory access;

data restrictions;

interoperability restrictions;

preferential ranking;

technical degradation;

exclusion from auctions;

discriminatory measurement;

self-preferencing.

The Google AdSense litigation demonstrates the importance of proving actual or potential exclusionary effects with sufficient evidentiary precision. (curia)

27. DMA and Transparency of Advertising Markets

The DMA introduces an important transparency mechanism.

Advertisers and publishers can request information concerning:

prices;

fees;

remuneration;

metrics;

advertising performance.

The DMA also provides access to certain performance-measurement tools and data so that business users can independently verify advertising inventory. (Digital Markets Act (DMA))

This addresses a fundamental information asymmetry:

If one platform controls both the advertising transaction and the information used to evaluate that transaction, independent verification becomes difficult.

28. Attention Allocation and Consumer Choice

Attention allocation can also affect consumers.

A platform controls:

what is shown;

what is recommended;

how advertisements are presented;

whether commercial content is distinguishable;

how easily consumers can reach alternatives.

The competition issue therefore overlaps with:

consumer protection;

transparency;

data protection;

platform regulation.

But these are legally distinct areas.

29. Role of GDPR

GDPR becomes especially important where attention allocation relies on profiling.

Potential sequence:

Personal data

→ profiling

→ personalisation

→ advertising targeting

→ attention

→ commercial conversion

The Meta judgment establishes that a competition authority may, in an abuse-of-dominance investigation, consider whether data processing complies with GDPR, while coordinating with the competent data-protection authorities. (curia)

30. Remedies

Possible European remedies include:

Behavioural remedies

non-discrimination;

transparent ranking;

equal access;

data-access obligations;

interoperability.

Structural remedies

In exceptional circumstances:

divestiture;

separation of business units;

separation of conflicting functions.

DMA remedies

The DMA provides specific compliance obligations and enforcement mechanisms for designated gatekeepers.

Financial penalties

Competition authorities can impose significant fines.

For example, the Commission's September 2025 Google ad-tech decision imposed a €2.95 billion fine. (European Commission)

31. Civil Liability and Private Actions

Although the principal authorities are competition regulators, private parties may potentially seek:

damages;

compensation;

contractual remedies;

injunctions;

restitution.

Potential claimants include:

advertisers;

publishers;

ad-tech intermediaries;

app developers;

competing platforms.

A claimant would generally need to establish:

Competition-law infringement → legally relevant harm → causation → quantifiable damage

depending on the applicable national procedural and damages framework.

32. Evidence in Ad-Tech Litigation

Ad-tech disputes are highly evidence-intensive.

Important evidence includes:

Algorithmic evidence

ranking algorithms;

recommendation models;

auction rules;

source code where obtainable;

A/B testing.

Commercial evidence

contracts;

exclusivity agreements;

pricing;

commissions;

auction data.

Technical evidence

APIs;

logs;

data flows;

interoperability documentation.

Data evidence

user identifiers;

audience segments;

tracking information;

cross-platform data.

Economic evidence

market shares;

diversion ratios;

foreclosure rates;

entry barriers;

switching costs;

advertiser behaviour.

33. Economic Test

A useful competition-law model is:

Market Power

  •  

Control of Attention/Data

  •  

Vertical Integration

  •  

Exclusionary Conduct

  •  

Foreclosure / Competitive Harm

=

Potential Abuse

The final step still requires application of the relevant legal standard.

34. Key Case-Law Table

CaseCourt/AuthorityCore principleAd-tech relevance
Google Shopping, C-48/22 PCJEUSelf-preferencing and leveraging through search visibility can constitute abuseRanking and attention allocation
Google Android, T-604/18General CourtEcosystem, tying, exclusivity and cross-market foreclosureEcosystem concentration
Google AdSense, T-334/19General CourtExclusive advertising-intermediation restrictions require rigorous proof of exclusionary effectsDirect ad-tech authority
Meta Platforms, C-252/21CJEUCompetition authorities may consider GDPR compliance in dominance casesData-driven advertising
Amazon, B2-88/18BundeskartellamtPlatform dependence and seller access can raise abuse concernsMarketplace visibility
Amazon, Section 19a proceedingsBundeskartellamtEcosystem power can justify enhanced scrutiny of platform conductCross-market concentration
Apple ATTF proceedingsBundeskartellamtDifferential tracking rules can raise self-preferencing concernsAdvertising data and targeting
Google Ad-tech, AT.40670European CommissionFavouring own online display-advertising technology services can infringe Article 102Direct ad-tech ecosystem concentration

The Google AdSense judgment is presently under appeal in C-826/24 P, so its final appellate status should be kept separate from the final CJEU judgments such as Google Shopping and Meta. (Infocuria)

35. Emerging Legal Issue: Control of "Attention Infrastructure"

The most important future issue is likely to be the emergence of attention infrastructure as a competitive asset.

Traditionally:

Market power = control over supply.

Increasingly:

Digital market power = control over access, data, ranking, recommendation and attention.

A platform controlling the interface through which consumers discover products can potentially influence several downstream markets simultaneously.

This is why the European regulatory approach increasingly examines ecosystems rather than isolated products.

The Commission's current DMA work also illustrates this trend: in 2026 it opened proceedings concerning Google's access to search ranking, query, click and view data for third-party search providers, including questions concerning access by AI-chatbot providers. (Digital Markets Act (DMA))

36. Legal Test for Ad-Tech Ecosystem Concentration

A comprehensive European analysis can use the following test:

Step 1 — Define the ecosystem

Identify:

search;

social;

video;

operating system;

advertising;

data;

analytics.

Step 2 — Identify the relevant markets

Consider:

advertiser side;

publisher side;

consumer side;

ad-intermediation side;

data-related services.

Step 3 — Measure ecosystem power

Examine:

market shares;

users;

data;

network effects;

switching costs;

multi-homing.

Step 4 — Identify the attention bottleneck

Determine who controls:

ranking;

recommendation;

search;

advertising placement;

default settings.

Step 5 — Examine vertical integration

Determine whether the undertaking operates at multiple levels.

Step 6 — Identify conduct

Look for:

self-preferencing;

exclusivity;

tying;

discrimination;

data restrictions;

interoperability restrictions.

Step 7 — Establish competitive effects

Examine:

foreclosure;

entry barriers;

reduced innovation;

reduced advertiser choice;

reduced publisher choice.

Step 8 — Consider objective justification

Not every restriction is unlawful.

Step 9 — Apply DMA where applicable

Determine whether a gatekeeper obligation applies independently of traditional dominance analysis.

Step 10 — Determine remedy

Possible remedies include:

transparency;

access;

interoperability;

non-discrimination;

behavioural commitments;

fines;

structural measures in exceptional circumstances.

37. Conclusion

Ad-Tech Ecosystem Concentration and Attention Allocation Control represents a modern European competition-law problem in which data, algorithms, interfaces, advertising infrastructure and user attention become interconnected sources of market power.

The most important legal concepts are:

Ecosystem Power → Data Advantage → Attention Control → Vertical Integration → Self-Preferencing/Exclusion → Foreclosure → Competition Harm

The case law shows several complementary principles:

Google Shopping demonstrates the importance of preferential ranking and visibility. (curia)

Google Android demonstrates how ecosystem integration can reinforce dominance across interconnected markets. (curia)

Google AdSense demonstrates the special complexity of advertising intermediation and the need for rigorous proof of exclusionary effects. (Infocuria)

Meta demonstrates the interaction between data protection and competition in an advertising-funded social-network ecosystem. (curia)

Amazon demonstrates how platform dependence and visibility rules can affect businesses operating inside a digital ecosystem. (Federal Cartel Office)

Apple's ATTF proceeding illustrates the emerging issue of whether a platform can impose advertising-data restrictions on rivals while treating its own ecosystem differently. (Federal Cartel Office)

The 2025 Google ad-tech decision shows that the EU is now addressing the advertising stack itself, rather than looking only at consumer-facing search or social platforms. (European Commission)

Core formula

Ad-Tech Ecosystem Concentration + Control of Data/Ranking/Advertising Infrastructure + Ability to Allocate User Attention + Exclusionary or Self-Preferential Conduct + Competitive Harm = Potential European Competition-Law Liability.

LEAVE A COMMENT