Digital Persuasion Systems And Platform Dominance Risks

Introduction

Digital persuasion systems are technological systems through which digital platforms influence, steer, or manipulate the choices of users, sellers, advertisers, workers, or business partners. They include recommendation algorithms, personalized feeds, behavioural advertising, ranking systems, dark patterns, default settings, notifications, nudges, personalized pricing, recommender systems, and AI-generated persuasive content.

The competition-law concern arises when a platform possessing substantial market power uses these systems not merely to compete on the merits but to reinforce its own dominance, exclude rivals, raise switching costs, exploit dependent users, or extend power into adjacent markets.

The central question is therefore:

When does legitimate digital persuasion become an instrument of exclusionary or exploitative market power?

Digital persuasion is particularly important because the platform can observe enormous quantities of behavioural data and continuously modify the interface in response to user behaviour. This can create a feedback loop:

Data → Personalisation → Behavioural influence → More engagement → More data → Stronger optimisation → Greater platform power

1. Meaning of Digital Persuasion Systems

A digital persuasion system is an automated or semi-automated mechanism designed to influence behaviour through digital interfaces or algorithms.

Major forms

A. Recommendation systems

Platforms determine which products, videos, applications, news items, or services users see.

Examples include:

• video recommendations;
• search rankings;
• product recommendations;
• social-media feeds;
• music recommendations;
• app rankings.

A dominant platform may theoretically manipulate recommendations to favour its own services.

B. Personalised advertising

Platforms use behavioural data to determine:

• which advertisements users receive;
• when advertisements appear;
• which products are promoted;
• which consumers are targeted;
• what price advertisers pay for particular audiences.

The competition concern increases when the platform controls both the audience data and the advertising infrastructure.

C. Dark patterns

Dark patterns are interface designs that steer users toward decisions they might not otherwise make.

Examples include:

• difficult cancellation procedures;
• pre-selected options;
• misleading consent buttons;
• repeated prompts;
• hidden alternatives;
• subscription traps.

D. Default settings

A platform may determine the default:

• search engine;
• browser;
• payment mechanism;
• app store;
• cloud service;
• digital assistant;
• advertising technology;
• identity provider.

Defaults can be extraordinarily powerful because many users never change them.

E. Ranking and visibility systems

A platform can control commercial visibility through:

• search ranking;
• seller ranking;
• app-store ranking;
• recommendation placement;
• preferred-provider badges;
• sponsored positioning.

For dependent businesses, visibility itself can become an essential commercial input.

F. AI persuasion

Generative AI and conversational assistants create a newer form of persuasion.

An AI system can:

• personalise recommendations;
• anticipate consumer preferences;
• selectively present information;
• recommend the platform's own services;
• negotiate or rank products;
• influence purchasing decisions;
• automate behavioural experimentation.

This creates a potential transition from platform as intermediary to platform as behavioural gatekeeper.

2. Why Digital Persuasion Creates Competition Risks

Traditional competition analysis frequently focuses on prices, output, market shares and costs.

Digital persuasion creates additional dimensions of competitive harm.

2.1 Attention control

A dominant platform can control scarce user attention.

If the platform decides which competing service appears first, competitors may technically remain available while becoming commercially invisible.

Thus:

Availability ≠ Effective access

2.2 Data advantage

A dominant platform may possess:

• search history;
• purchase history;
• location information;
• engagement data;
• click-through behaviour;
• transaction data;
• advertising-response data.

The resulting behavioural dataset can make the platform's persuasion technology more effective than that of smaller competitors.

2.3 Network effects

The more users a platform has, the more behavioural information it generates.

More data can improve recommendations, which can attract more users, producing a reinforcing cycle.

Users → Data → Better prediction → Better persuasion → More users

2.4 Switching costs

Personalisation can create invisible switching costs.

Users may have:

• personalised histories;
• saved preferences;
• recommendations;
• social graphs;
• accumulated reputation;
• loyalty benefits;
• stored payment credentials.

Consequently, a competing platform may need to overcome not only price differences but an entire personalised digital environment.

3. Digital Persuasion and Article 102 TFEU / UK Competition Law

Under European competition law, Article 102 TFEU prohibits the abuse of a dominant position where conduct affects trade between Member States.

Relevant theories include:

• exclusionary abuse;
• tying;
• bundling;
• discriminatory treatment;
• refusal of access;
• self-preferencing;
• exploitative conduct;
• leveraging dominance into adjacent markets.

In the UK, comparable analysis principally arises under Chapter II of the Competition Act 1998, together with the newer digital-markets framework.

The important analytical shift is that conduct may be problematic even when the immediate consumer price is zero.

The relevant competitive variables can instead include:

• attention;
• data;
• quality;
• privacy;
• innovation;
• interoperability;
• choice;
• visibility;
• switching costs.

4. Digital Persuasion as a Form of Self-Preferencing

Suppose a dominant marketplace operates both:

1. the marketplace infrastructure; and
2. its own competing retail products.

Its algorithm could rank its own products more prominently.

The system may formally claim to optimise:

"customer relevance."

But the competition-law question is whether the algorithm actually operates to foreclose competitors.

The problem becomes particularly serious when the platform controls the principal route through which consumers discover competing products.

5. The Feedback-Loop Problem

One of the most important risks is algorithmic reinforcement.

Suppose a platform promotes its own product.

1. Users see the product more frequently.
2. Users purchase it more frequently.
3. The platform records greater engagement.
4. The algorithm interprets engagement as evidence of relevance.
5. The product receives still greater visibility.
6. Rival products receive less exposure.
7. Rival sales decline.
8. The platform's own product becomes increasingly dominant.

This produces:

Algorithmic preference → behavioural response → data signal → algorithmic reinforcement

The critical legal problem is that apparently neutral algorithmic optimisation can potentially reproduce an initial exclusionary preference.

6. Important Case Laws

1. Google Search (Shopping) — European Commission / General Court

The Google Shopping litigation is one of the most important authorities for digital-platform self-preferencing.

Google operated the dominant general-search service while also providing its own comparison-shopping service.

The European Commission found that Google systematically favoured its comparison-shopping service in search-result positioning while demoting competing comparison-shopping services.

The General Court substantially upheld the Commission's findings.

Relevance to digital persuasion

The case demonstrates that competition law can examine ranking and visibility, rather than merely monetary pricing.

A platform can potentially distort competition by controlling the algorithmic pathway through which users discover competing services.

Principle

Control over user attention and ranking can constitute a significant competitive advantage when exercised by a dominant platform.

2. Google Android — European Commission / General Court

The Android litigation concerned Google's practices involving mobile operating systems, search, browsers and app distribution.

The Commission examined arrangements including:

• pre-installation;
• search-related requirements;
• revenue-sharing arrangements;
• restrictions concerning Android forks.

Relevance

Pre-installation and default positioning can operate as digital persuasion mechanisms.

A consumer may select the pre-installed service simply because it is:

• immediately available;
• familiar;
• integrated;
• difficult to replace.

Thus, the competitive significance of defaults can exceed their apparent technical simplicity.

3. Google Android Auto — European Commission

The Android Auto case involved Google's refusal to provide interoperability for certain applications.

Although not a classic "persuasion" case, it is relevant because digital ecosystems can determine which applications users can meaningfully access.

Competition significance

A platform can influence consumer behaviour not only by recommending one product but also by determining whether competing services can participate in the ecosystem at all.

This creates a continuum:

Visibility → Ranking → Default → Interoperability → Access

4. Amazon Marketplace — European Commission

The European Commission's Amazon investigations concerned the use of non-public marketplace seller data and Amazon's potential role as both marketplace operator and competing retailer.

The underlying concern illustrates the importance of dual-role platforms.

Amazon potentially possesses information generated by third-party sellers while simultaneously competing against those sellers.

Relevance to persuasion

A platform possessing detailed seller and consumer information may theoretically optimise:

• product positioning;
• advertising;
• recommendation;
• search ranking;
• product development.

The competitive danger is therefore not simply data collection but data-enabled strategic influence.

5. British Airways v Commission

In British Airways v Commission, the Court of Justice examined loyalty-related arrangements operated by a dominant undertaking.

The case concerned incentive structures that could encourage travel agents to favour British Airways.

Digital relevance

Although the case predates modern digital platforms, its economic logic is highly relevant to digital persuasion.

A digital platform may similarly use:

• loyalty programmes;
• personalised rewards;
• ranking incentives;
• seller discounts;
• preferential visibility;
• algorithmically targeted benefits.

The essential question is whether incentives foreclose competitors rather than merely rewarding legitimate efficiency.

6. Intel v Commission

The Intel litigation concerned rebates granted by a dominant undertaking.

The modern significance of the case lies in the requirement to examine whether allegedly exclusionary rebates are capable of foreclosing an equally efficient competitor.

Digital relevance

A digital platform could theoretically combine:

• personalised discounts;
• targeted incentives;
• advertising credits;
• preferential placement;
• loyalty benefits.

These mechanisms may be individually small but collectively powerful.

Competition analysis therefore needs to examine their economic effects, rather than treating each digital nudge as an isolated event.

7. United States v Google

The US Google search litigation provides another important framework for understanding digital persuasion.

The case examined Google's agreements and distribution practices relating to default search placement.

Digital persuasion significance

Search defaults affect consumer behaviour because users often rely on the service already presented by:

• browsers;
• mobile devices;
• operating systems;
• other distribution channels.

Thus, control of the choice architecture can reinforce search-market dominance.

The broader principle is that competition can be harmed where a dominant firm uses distribution arrangements to make alternative services substantially harder to reach.

8. Epic Games v Apple

The Epic Games litigation concerning Apple's App Store provides an important example of competition issues arising from platform governance.

The dispute involved:

• app distribution;
• payment systems;
• platform rules;
• commissions;
• restrictions on alternative payment mechanisms.

Relevance

A platform's ability to design the environment in which businesses interact with consumers gives it substantial behavioural and commercial gatekeeping power.

The platform can determine:

• what consumers see;
• what transactions are permitted;
• which payment mechanisms are available;
• how developers communicate with customers.

Digital persuasion therefore operates alongside platform rule-making power.

7. Persuasion and Consumer Choice

Competition law must distinguish between persuasion and coercion.

Advertising is inherently persuasive.

A platform recommending a product is not automatically anticompetitive.

The legal concern becomes stronger where several factors coincide:

1. dominance;
2. dependence of users or businesses;
3. lack of effective alternatives;
4. opaque ranking mechanisms;
5. self-preferencing;
6. discriminatory access;
7. exclusionary incentives;
8. high switching costs;
9. exploitation of behavioural data;
10. measurable foreclosure.

8. Dark Patterns and Competition Law

Dark patterns traditionally belong partly to consumer-protection law.

However, they may acquire competition significance where a dominant platform uses them to:

• prevent users from switching;
• discourage multi-homing;
• force acceptance of platform services;
• favour its own ecosystem;
• make rival services difficult to access.

For example:

Easy registration → difficult cancellation → reduced switching → greater retention → stronger network effects

The competition issue is therefore not merely that users are manipulated.

It is whether manipulation protects or extends market power.

9. Personalisation and Price Discrimination

AI systems can potentially personalise prices according to predicted willingness to pay.

For example:

• Consumer A receives one offer.
• Consumer B receives another.
• Consumer C receives a different promotion.

Personalisation itself is not automatically unlawful.

However, competition concerns can arise where a dominant platform:

• discriminates against customers of competing services;
• uses rival-specific data;
• rewards ecosystem loyalty;
• increases switching costs;
• forecloses competing platforms.

10. Behavioural Data as a Strategic Asset

Traditional market power analysis frequently examines financial assets, infrastructure and intellectual property.

Digital persuasion introduces another strategic asset:

Predictive behavioural knowledge.

A dominant platform may know not merely what consumers bought but:

• what they considered;
• what they rejected;
• how long they hesitated;
• what they searched for;
• which recommendation changed their behaviour;
• what advertisement triggered engagement.

This can produce an information asymmetry between the platform and its users or competitors.

11. Persuasion, Multi-Homing and Lock-In

Competition is stronger when users can easily multi-home.

If consumers routinely use:

• Google and Bing;
• Amazon and other marketplaces;
• several social networks;
• multiple payment services,

platform dominance is potentially constrained.

Persuasive design can weaken multi-homing by making the dominant platform increasingly convenient.

Examples include:

• synchronised accounts;
• personalised recommendations;
• exclusive rewards;
• integrated payments;
• automatic login;
• cross-service identity;
• ecosystem-specific benefits.

This creates:

Persuasion → Reduced multi-homing → Greater network effects → Greater dominance

12. AI Agents Create a New Dimension

AI agents may become autonomous purchasing or service-selection intermediaries.

Instead of a consumer manually searching:

"Find me a hotel."

an AI agent may decide:

"This is the best hotel for you."

The competitive issue becomes extremely important if the AI agent is controlled by a dominant platform.

The platform could potentially influence:

• which suppliers are considered;
• which products are ranked;
• which offers are displayed;
• which suppliers receive recommendations;
• what information is omitted.

This could shift competition from:

"Which product does the consumer choose?"

to:

"Which products does the algorithm allow the consumer to consider?"

13. Algorithmic Choice Architecture

Digital platforms increasingly construct the environment in which choices occur.

A simplified hierarchy is:

Level 1 — Information

What information does the user receive?

Level 2 — Ranking

What appears first?

Level 3 — Recommendation

What is actively suggested?

Level 4 — Default

What happens automatically?

Level 5 — Friction

How difficult is it to choose an alternative?

Level 6 — Lock-in

How difficult is it to leave?

The higher the platform moves along this hierarchy, the greater the potential competition concern.

14. Legal Test for Digital Persuasion Abuse

A useful competition-law framework is:

Step 1 — Define the relevant market

Determine whether the platform operates in:

• search;
• social networking;
• digital advertising;
• marketplaces;
• app distribution;
• cloud;
• payments;
• AI services.

Step 2 — Establish dominance

Examine:

• market share;
• network effects;
• switching costs;
• entry barriers;
• data advantages;
• ecosystem integration;
• user dependence.

Step 3 — Identify the persuasion mechanism

Ask whether the platform uses:

• ranking;
• recommendations;
• defaults;
• incentives;
• personalised offers;
• interface manipulation;
• data-driven targeting.

Step 4 — Identify the competitive effect

Determine whether the conduct:

• excludes rivals;
• raises rivals' costs;
• reduces multi-homing;
• increases switching costs;
• forecloses distribution;
• disadvantages dependent businesses.

Step 5 — Examine efficiencies

The platform may argue that the system produces:

• better relevance;
• improved security;
• lower transaction costs;
• fraud prevention;
• improved user experience;
• innovation.

Step 6 — Proportionality and remedy

Authorities must consider whether less restrictive alternatives exist.

15. Possible Remedies

Competition authorities could consider:

A. Ranking transparency

Platforms could be required to disclose the principal parameters governing rankings.

B. Choice screens

Users may be presented with genuine alternatives rather than a single default.

C. Data portability

Users should be able to transfer relevant data to competing services.

D. Interoperability

Competing services may need technical access to essential ecosystem functions.

E. Non-discrimination

Dominant platforms could be prohibited from systematically disadvantaging rivals.

F. Restrictions on self-preferencing

The platform may be required to apply comparable ranking criteria to its own and third-party services.

G. Interface neutrality

Certain manipulative interface practices could be prohibited.

H. Algorithmic auditing

Regulators may examine whether ranking or recommendation systems systematically produce exclusionary outcomes.

16. Key Distinction: Consumer Protection vs Competition Law

Not every manipulative interface is a competition violation.

The same conduct can therefore trigger multiple legal regimes simultaneously.

17. Emerging Doctrine: From Market Power to Behavioural Power

Digital platforms demonstrate that market power is no longer limited to the ability to raise prices.

A platform can possess:

Behavioural power — the ability to systematically influence how market participants perceive, evaluate and select alternatives.

This creates a broader conception of dominance:

Economic power + informational power + infrastructural power + behavioural power

A platform that controls all four may be capable of shaping the competitive structure itself.

Conclusion

Digital persuasion systems represent an important frontier of competition law because they allow dominant platforms to influence markets through attention, ranking, defaults, personalisation, data and behavioural design.

The principal danger is not persuasion by itself. Competition law must distinguish ordinary commercial persuasion from situations in which a dominant platform uses its gatekeeper position to manipulate the competitive process.

The most significant risks are:

1. self-preferencing;
2. algorithmic ranking discrimination;
3. default-based exclusion;
4. reduced multi-homing;
5. behavioural lock-in;
6. data-enabled foreclosure;
7. personalised exclusionary incentives;
8. dark-pattern-enabled switching barriers;
9. AI-mediated recommendation bias; and
10. conversion of behavioural influence into durable market dominance.

The emerging legal principle can therefore be expressed as:

Where a dominant digital platform controls both the market interface and the mechanisms through which users make choices, algorithmic persuasion can become a competition-law instrument capable of reinforcing or extending market power.