Influence-As-A-Service Ecosystems And Behavioural Manipulation Risks

1. Introduction

Influence-as-a-Service (IaaS) refers to commercial ecosystems in which firms provide technology, data, algorithms, behavioural analytics, targeting infrastructure, content-distribution tools, recommendation systems, reputation management, advertising optimisation, influencer networks, or automated persuasion services to influence the behaviour of individuals or groups.

The concept extends beyond conventional advertising. An influence-as-a-service provider may combine:

• behavioural profiling;
• real-time attention and engagement analytics;
• personalised recommendations;
• targeted advertising;
• influencer or creator networks;
• automated content generation;
• social-media amplification;
• algorithmic ranking;
• sentiment and emotion analysis;
• dynamic pricing or personalised offers;
• political or public-affairs messaging;
• reputation-management services; and
• AI systems that optimise persuasive outcomes.

From a competition-law perspective, the central concern is not simply that consumers are persuaded. Persuasion becomes legally significant where a powerful platform or intermediary uses market power, data advantages, algorithmic control, contractual restrictions, or ecosystem integration to manipulate behavioural choice, exclude rivals, exploit users, or distort competitive conditions.

2. Meaning of Influence-as-a-Service

An Influence-as-a-Service ecosystem can be represented as:

Data collection → Profiling → Prediction → Targeting → Content generation → Distribution → Behavioural response → Measurement → Further optimisation

For example:

A platform collects behavioural data → predicts that a consumer is highly price-sensitive → generates a personalised advertisement → determines the optimal time and ranking for displaying it → directs the consumer toward a preferred merchant → measures whether the consumer purchased → feeds that result back into the model.

The service provider therefore sells not merely advertising space, but optimised behavioural influence.

This creates a potentially important distinction:

3. Competition-Law Relevance

Influence-as-a-Service can create competition concerns under several theories.

A. Abuse of dominance

A dominant digital platform may use control over:

• user data;
• advertising inventory;
• search rankings;
• app distribution;
• social graphs;
• recommendation systems;
• identity systems; or
• measurement infrastructure

to disadvantage competing influence providers.

B. Exploitative conduct

The platform may exploit consumers through:

• excessive personalisation;
• dark patterns;
• behavioural targeting;
• opaque recommendation systems;
• discriminatory offers; or
• manipulation of attention.

C. Foreclosure

A dominant platform could prevent advertisers or influence providers from using competing services by imposing:

• exclusivity;
• tying;
• self-preferencing;
• data-access restrictions;
• interoperability restrictions;
• anti-steering provisions; or
• restrictions on multi-homing.

D. Data-driven market power

The competitive advantage may arise from the combination of:

data + computing power + behavioural prediction + distribution + feedback loops.

The resulting advantage may be difficult for rivals to reproduce.

4. Behavioural Manipulation as a Competition Concern

Not every persuasive technique constitutes an antitrust violation.

The important question is whether behavioural manipulation is connected to competitive harm or abuse of market power.

A useful analytical model is:

Market Power → Behavioural Data Advantage → Algorithmic Influence → Reduced Choice → Competitive Harm

Potential competitive effects include:

1. reduced consumer autonomy;
2. switching suppression;
3. increased consumer lock-in;
4. exclusion of rival suppliers;
5. reduced transparency;
6. discriminatory access to consumers;
7. exploitation of captive users;
8. degradation of quality or privacy;
9. manipulation of demand;
10. weakening of competitive entry.

5. Influence-As-a-Service and Network Effects

Influence platforms often benefit from powerful network effects.

More users generate more:

• behavioural data;
• social connections;
• engagement information;
• conversion data;
• content;
• advertiser interest.

This improves the platform's ability to predict and influence behaviour.

The cycle becomes:

More users → More data → Better prediction → Better targeting → More advertisers → More revenue → More investment → More users

This can produce a data-feedback network effect.

A new entrant may therefore face a significant competitive disadvantage even if it possesses technically comparable algorithms.

6. The Role of Artificial Intelligence

AI substantially increases the scale of influence-as-a-service.

AI systems can automatically:

• classify users;
• infer preferences;
• predict purchasing behaviour;
• identify emotional states;
• generate personalised messages;
• determine optimal communication timing;
• test thousands of variants;
• optimise engagement;
• identify susceptible audiences;
• automate influencer campaigns; and
• continuously modify strategies.

The concern is particularly acute where the optimisation function is not:

"Provide useful information."

but:

"Maximise behavioural conversion."

An algorithm optimised exclusively for engagement could favour content that provokes anger, fear, outrage, compulsive consumption, or polarisation because those responses may increase engagement.

7. Dark Patterns and Behavioural Steering

Influence-as-a-service can overlap with dark patterns.

Examples include:

• misleading interface design;
• hidden subscription renewal;
• difficult cancellation;
• default selections;
• countdown pressure;
• personalised scarcity messages;
• repeated prompts;
• social-pressure notifications;
• emotionally targeted recommendations.

Competition authorities may become interested where these practices are employed by dominant platforms to extract value from users or suppress competitive alternatives.

8. Self-Preferencing Through Influence Infrastructure

A dominant platform may provide influence services to businesses while simultaneously operating a marketplace.

This creates a structural conflict.

For example:

Platform

→ owns advertising system
→ owns marketplace
→ owns consumer data
→ controls ranking
→ sells influence optimisation
→ competes with merchants using the platform.

The platform may potentially use its informational advantage to favour its own products or preferred sellers.

The issue therefore moves beyond traditional advertising into control over commercial visibility.

9. Behavioural Lock-In

Influence systems can reinforce switching costs.

A platform may learn:

• what a consumer purchases;
• which content attracts attention;
• preferred brands;
• price sensitivity;
• browsing patterns;
• social connections;
• responsiveness to particular messages.

A competing platform lacks equivalent information.

Consequently, consumers may remain within the incumbent ecosystem because the incumbent provides increasingly personalised experiences.

This creates:

Personalisation → Dependency → Switching Cost → Reduced Contestability

10. Six Important Case Laws

Case 1: Google Shopping — European Commission

The Google Shopping decision is highly relevant to influence-as-a-service because it concerns the use of a dominant search platform's infrastructure to favour its own comparison-shopping service.

The European Commission found that Google systematically gave prominent placement to its own comparison-shopping service while demoting competing services.

Relevance

The case demonstrates that control over visibility and ranking can constitute an important competitive advantage.

For Influence-as-a-Service ecosystems, the analogous concern is:

A dominant platform controls the algorithm determining who receives behavioural attention.

If the platform systematically directs users toward its own commercial services or selected partners, the influence mechanism can become a form of exclusionary conduct.

Principle

Control over digital visibility can constitute an instrument of market power.

11. Case 2: Facebook/Meta — German Federal Cartel Office

The German Facebook proceedings concerning extensive collection and combination of user data are particularly important.

The Bundeskartellamt examined Facebook's ability to combine data obtained from Facebook with information from other services and third-party websites.

The case established an important connection between:

data practices + dominance + user autonomy + competition law.

Relevance to Influence-as-a-Service

An influence platform with access to enormous quantities of behavioural information can create a competitive advantage unavailable to smaller rivals.

The greater the quantity and diversity of behavioural information, the greater the potential ability to:

• predict behaviour;
• target users;
• personalise content;
• optimise advertising; and
• increase user dependency.

Principle

Data accumulation can become a competition concern when connected with market power and exploitative conditions.

12. Case 3: FTC v Facebook / Meta

The US antitrust litigation concerning Facebook's acquisitions and competitive strategy provides another important framework.

The Federal Trade Commission alleged that Facebook maintained monopoly power in personal social networking services and engaged in strategies that impeded competitive threats.

Although the litigation involves questions distinct from behavioural manipulation, it is relevant because social-network dominance can give an undertaking control over an extraordinarily valuable attention and influence infrastructure.

Relevance

A dominant social platform can potentially control:

• audience access;
• social distribution;
• advertising infrastructure;
• behavioural information;
• recommendation systems.

This creates a powerful ecosystem in which influence services can become dependent on the incumbent.

Principle

Control over an important digital ecosystem can create competitive advantages extending beyond the immediate service.

13. Case 4: FTC v Amazon

The US Federal Trade Commission's Amazon litigation is relevant to behavioural influence because it addresses alleged practices involving Amazon's marketplace, sellers, pricing, advertising and consumer-facing ecosystem.

Amazon simultaneously operates as:

• marketplace;
• retailer;
• advertising provider;
• logistics provider;
• data intermediary;
• recommendation system.

Relevance

This demonstrates the potential competition problem created when a platform controls both:

the commercial environment + the tools used to influence commercial behaviour within that environment.

A dominant marketplace can potentially influence:

• which products consumers see;
• which sellers receive visibility;
• how products are ranked;
• how advertising interacts with organic placement.

Principle

Vertical integration of marketplace infrastructure and behavioural influence tools can raise foreclosure and self-preferencing concerns.

14. Case 5: Commission v Google — Android

The Google Android decision concerned Google's contractual arrangements involving mobile-device manufacturers and the Android ecosystem.

The case illustrates how control over an important digital access point can reinforce dominance in adjacent markets.

Relevance to Influence-as-a-Service

Mobile devices provide the gateway through which users encounter:

• search;
• applications;
• advertising;
• notifications;
• recommendations;
• digital assistants.

Control over defaults can therefore influence consumer behaviour at the earliest stage of the digital journey.

Principle

Control over access points and defaults can reinforce dominance in downstream digital services.

15. Case 6: Google AdSense

The European Commission's AdSense decision concerned Google's contractual restrictions relating to search advertising intermediaries on third-party websites.

The Commission found that Google's contractual restrictions prevented rivals from competing effectively in the market for online search advertising intermediation.

Relevance

AdSense is especially relevant because Influence-as-a-Service frequently operates through advertising intermediaries.

A dominant provider controlling:

advertiser → intermediary → publisher → consumer

may influence the entire chain.

Restrictive contractual arrangements can therefore prevent competing influence providers from gaining sufficient scale.

Principle

A dominant intermediary can potentially use contractual restrictions to protect its position across an influence ecosystem.

16. Case 7: Apple App Store / Epic Games

The Epic Games litigation concerning Apple's App Store ecosystem is highly relevant to digital influence infrastructure.

Apple controls:

• app distribution;
• payment infrastructure;
• platform rules;
• access to consumers.

Epic challenged Apple's restrictions concerning alternative payment mechanisms and distribution.

Relevance

An influence provider operating through applications may become dependent upon the platform's:

• distribution rules;
• payment mechanisms;
• ranking systems;
• access conditions.

The broader competition issue is platform dependency.

Principle

Control over a digital gateway can give the platform substantial leverage over downstream businesses.

17. Case 8: Booking.com and Online Intermediation

The European competition-law litigation and enforcement concerning online booking platforms has also contributed to the development of principles concerning platform intermediation, parity clauses and consumer choice.

Such cases are relevant because platforms can influence consumer decisions through:

• rankings;
• recommendations;
• visibility;
• personalised presentation;
• contractual restrictions.

The competitive question is whether platform mechanisms merely facilitate transactions or shape the competitive environment in favour of the platform or selected participants.

18. Influence-As-a-Service and Self-Preferencing

A particularly serious risk arises where the provider of influence infrastructure also competes with its customers.

Consider:

Platform owns data + platform owns recommendation algorithm + platform provides advertising + platform sells its own products.

The platform could potentially obtain information about:

• emerging consumer demand;
• competitors' conversion rates;
• successful campaigns;
• price sensitivity;
• product popularity.

That information can potentially be used competitively against downstream customers.

This produces a dual-role problem.

19. Algorithmic Collusion Risk

Influence-as-a-Service may also facilitate coordination among competitors.

Suppose several competing firms purchase the same algorithmic optimisation service.

The provider may have access to:

• prices;
• inventory;
• demand;
• conversion rates;
• competitor behaviour;
• campaign strategies.

If algorithms use shared information to optimise outcomes, the system could potentially facilitate coordinated behaviour without traditional human communication.

Competition authorities therefore need to distinguish between:

independent algorithmic optimisation

and

algorithmically facilitated coordination.

20. Personalised Pricing and Behavioural Exploitation

Influence systems can potentially be connected to personalised pricing.

For example, algorithms could infer:

• willingness to pay;
• urgency;
• purchasing frequency;
• switching probability;
• brand loyalty.

A dominant firm could theoretically use this information to identify consumers who are unlikely to switch and charge them differently.

Competition law may become concerned where such conduct is connected to:

• dominance;
• exclusion;
• discrimination;
• exploitative abuse;
• consumer harm.

21. Attention as a Competitive Resource

Traditional competition analysis often concentrates on:

• price;
• output;
• market share.

Digital influence ecosystems introduce another resource:

consumer attention.

A platform competes not only for purchases but for:

• clicks;
• screen time;
• engagement;
• subscriptions;
• behavioural responses.

Where one firm controls a large portion of consumer attention, competitors may become dependent on that firm's infrastructure.

This can create an attention bottleneck.

22. Data-Driven Entry Barriers

New competitors may face three major barriers:

First — Data disadvantage

They lack historical behavioural information.

Second — Distribution disadvantage

They lack access to the incumbent's audience.

Third — Feedback disadvantage

They cannot collect enough interactions to improve their algorithms.

The resulting structure can be represented as:

Incumbent scale → data → better prediction → better influence → more users → more data

This can make the market increasingly difficult to contest.

23. Influence Infrastructure as an Essential Input

In extreme circumstances, an influence platform may become an indispensable gateway to consumers.

For example, if a dominant platform controls:

• search advertising;
• social advertising;
• app distribution;
• marketplace visibility;
• recommendation systems,

a rival may technically be able to enter the market but lack any commercially viable method of reaching consumers.

This raises questions analogous to essential-facilities and refusal-to-deal doctrines, although the stringent legal requirements for those doctrines remain important.

24. Consumer Welfare and Non-Price Harm

Influence-as-a-Service illustrates why digital competition analysis cannot focus exclusively on price.

Consumers may receive services for zero monetary price while suffering:

• loss of privacy;
• manipulation;
• reduced choice;
• behavioural dependency;
• reduced transparency;
• lower quality;
• reduced ability to switch.

Accordingly, competition authorities increasingly need to consider quality, privacy, autonomy and innovation alongside price.

25. Relationship with Consumer Protection Law

Competition law and consumer protection law can overlap.

A behavioural manipulation practice might simultaneously involve:

Competition law

→ dominance / exclusion / foreclosure

Consumer law

→ deception / unfair practices / dark patterns

Data protection

→ unlawful processing / profiling / consent

Digital regulation

→ platform transparency / recommender-system obligations.

The same conduct can therefore produce multiple regulatory consequences.

26. Enforcement Difficulties

Influence-as-a-Service creates significant evidentiary problems.

A. Opaque algorithms

Authorities may not know why particular users received particular content.

B. Continuous experimentation

Algorithms may change thousands of times.

C. Individualised effects

Different consumers receive different treatment.

D. Difficult counterfactuals

Authorities must ask:

What would have happened without the algorithmic intervention?

E. Data asymmetry

The platform possesses information unavailable to regulators and competitors.

F. Attribution

It may be difficult to establish whether harmful behaviour resulted from:

• human instructions;
• algorithmic optimisation;
• training data;
• third-party advertisers;
• autonomous system decisions.

27. Appropriate Competition-Law Tests

A competition authority examining Influence-as-a-Service should consider:

1. Relevant market

Possible markets include:

• digital advertising;
• advertising intermediation;
• social-network services;
• recommendation services;
• behavioural analytics;
• influence optimisation;
• creator/influencer intermediation.

2. Market power

Assess:

• market share;
• data advantages;
• network effects;
• switching costs;
• multi-homing;
• access to attention;
• interoperability;
• entry barriers.

3. Conduct

Identify:

• self-preferencing;
• tying;
• exclusivity;
• discriminatory access;
• data leveraging;
• ranking manipulation;
• interoperability restrictions;
• anti-steering.

4. Effects

Assess:

• foreclosure;
• reduced innovation;
• reduced choice;
• exploitation;
• increased dependency;
• degradation of quality.

28. Remedies

Possible remedies include:

Structural remedies

• separation of advertising and marketplace operations;
• divestiture;
• functional separation.

Behavioural remedies

• non-discrimination;
• transparent ranking;
• prohibition of self-preferencing;
• data-access obligations.

Data remedies

• interoperability;
• data portability;
• restrictions on cross-service data combination.

Algorithmic remedies

• independent auditing;
• explainability requirements;
• testing of recommender systems;
• documentation of optimisation objectives.

Consumer remedies

• meaningful opt-outs;
• anti-dark-pattern requirements;
• clearer disclosures;
• limits on sensitive behavioural targeting.

29. Six Core Legal Principles Emerging from the Case Law

30. Hypothetical Example

Assume InfluenceCo operates the largest behavioural-targeting platform.

It provides:

1. consumer profiling;
2. AI-generated advertisements;
3. influencer selection;
4. recommendation placement;
5. behavioural prediction;
6. conversion optimisation.

InfluenceCo then acquires a competing advertising platform.

After the acquisition, it:

• restricts rival advertisers' access to behavioural data;
• gives its own influence tools superior access;
• requires advertisers to use its analytics system;
• bundles advertising with marketplace placement;
• uses marketplace data to improve its own targeting;
• prevents customers from exporting campaign-performance data.

The competition concerns would include:

Data leveraging + tying + self-preferencing + exclusion + interoperability restrictions + acquisition-related foreclosure.

The behavioural-manipulation element becomes especially important if InfluenceCo also optimises its system to maximise consumer dependency.

31. Distinguishing Legitimate Influence from Anticompetitive Manipulation

It is important not to treat all influence as unlawful.

Legitimate

• ordinary advertising;
• personalised recommendations;
• contextual advertising;
• consumer choice tools;
• loyalty programmes;
• product recommendations.

Potentially problematic

• concealed manipulation;
• discriminatory targeting;
• exploitation of vulnerable users;
• artificial suppression of rivals;
• self-preferencing;
• tying;
• exclusionary data practices;
• manipulation made possible by dominant-market control.

The decisive issue is therefore not influence itself, but the combination of influence technology and market power.

32. Future Competition-Law Challenge

AI may transform Influence-as-a-Service into autonomous influence optimisation.

An AI system could theoretically determine:

Who should be influenced → with what message → through which channel → at what time → at what price → with what emotional framing → based on which predicted vulnerability.

This raises a fundamental competition-law question:

Should market power include the ability to systematically shape consumer decision-making itself?

Traditional competition law generally protects the competitive process rather than guaranteeing completely autonomous consumer choice. However, where behavioural control is combined with dominance, exclusion, data exploitation, or reduced contestability, influence may become an important component of an abuse analysis.

33. Conclusion

Influence-as-a-Service ecosystems represent a new layer of digital market power. Their significance arises from the integration of data, AI, behavioural prediction, advertising, recommendation systems, social distribution and continuous optimisation.

The major competition risks are:

1. behavioural manipulation;
2. data-driven entry barriers;
3. platform dependency;
4. self-preferencing;
5. exclusion of rival influence providers;
6. algorithmic coordination;
7. personalised exploitation;
8. attention monopolisation;
9. reduced consumer choice; and
10. reinforcement of ecosystem dominance.

The most important doctrinal lesson from cases involving Google, Facebook/Meta, Amazon, Apple and digital intermediaries is that competition law increasingly has to examine not merely the price of a service, but who controls the infrastructure through which consumers discover, evaluate and ultimately choose products and services.

In an Influence-as-a-Service economy, control over behavioural information and consumer attention can become a form of competitive infrastructure. Where that infrastructure is controlled by a dominant undertaking and deployed to foreclose rivals or exploit users, it can present a significant modern competition-law problem.