AI Border Control Systems and Digital Market Partitioning

Introduction

AI border control systems are technologies used by governments and border authorities to automate or assist decisions concerning immigration, customs, security screening, identity verification, risk assessment, passenger processing and cross-border movement. They may include facial recognition, biometric databases, automated passport gates, predictive risk-scoring systems, AI-assisted customs inspection, passenger-information systems and algorithmic watch-list matching.

Digital market partitioning occurs when technological, contractual, regulatory or algorithmic arrangements divide a market into protected geographic, user, platform, data or service segments and thereby restrict effective competition across those segments.

The competition-law significance arises when a firm controlling an important digital infrastructure—such as an identity platform, biometric system, cloud environment, border-data network, travel platform or interoperability standard—uses that position to exclude competitors, discriminate between territories, restrict interoperability, foreclose cross-border supply or create artificial geographic barriers.

A useful analytical distinction is:

Legitimate border regulation is not itself an antitrust violation. Competition concerns arise where public regulatory requirements are supplemented by private conduct that unnecessarily partitions markets or where a dominant undertaking exploits border-control infrastructure to restrict competition.

I. Meaning and Structure of AI Border Control Systems

AI border-control systems generally contain several layers:

1. Identity layer

This includes:

• facial recognition;
• fingerprints;
• iris recognition;
• biometric passports;
• digital identity wallets;
• document authentication;
• identity matching.

2. Risk-assessment layer

Algorithms may evaluate:

• passenger information;
• travel histories;
• visa information;
• customs declarations;
• watch-list information;
• behavioural indicators;
• transaction or logistics information.

3. Decision-support layer

AI may recommend:

• secondary inspection;
• additional verification;
• customs examination;
• immigration referral;
• fraud investigation;
• risk classification.

4. Infrastructure layer

The system may depend upon:

• cloud infrastructure;
• government databases;
• API interfaces;
• identity providers;
• telecommunications networks;
• airport systems;
• airline reservation systems.

5. Interoperability layer

Competition becomes particularly significant where different national systems must communicate with:

• airlines;
• airports;
• customs agencies;
• immigration authorities;
• security agencies;
• travel platforms;
• digital identity providers.

II. What Is Digital Market Partitioning?

Digital market partitioning occurs when competitors or customers that could otherwise participate in the same competitive market are separated into different segments.

Partitioning can occur through:

A. Geographic restrictions

A digital service may work in:

• Country A but not Country B;
• one customs zone but not another;
• one airport but not competing airports.

B. Data localisation

A provider may insist that data generated in one territory cannot be processed elsewhere.

C. API restrictions

An incumbent can prevent competing providers from obtaining access to essential interfaces.

D. Interoperability restrictions

A dominant system may refuse to communicate with rival systems.

E. Algorithmic discrimination

Algorithms may provide different access, prices, rankings or functionality according to location.

F. Contractual territorial restrictions

Agreements may prohibit distributors or technology providers from serving customers outside assigned territories.

III. Competition-Law Theory

Several competition-law doctrines can potentially become relevant.

1. Abuse of Dominance

Where an AI border-control provider has substantial market power, discriminatory or exclusionary conduct may constitute abuse.

Relevant conduct can include:

• discriminatory access;
• refusal to supply;
• tying;
• exclusive dealing;
• self-preferencing;
• interoperability restrictions;
• discriminatory API access;
• excessive switching costs.

The key question is whether the conduct protects legitimate security requirements or instead exploits market power to exclude competitors.

2. Refusal to Deal

Suppose an airport operates an AI identity-verification infrastructure controlled by a dominant technology supplier.

If competing identity providers cannot realistically access the system, the incumbent may acquire an artificial competitive advantage.

The analysis would normally consider:

1. whether the infrastructure is indispensable;
2. whether duplication is realistically possible;
3. whether access has been refused;
4. whether the refusal eliminates effective competition;
5. whether objective justification exists.

IV. Essential-Facility Dimension

AI border infrastructure can theoretically acquire essential-facility characteristics.

For example:

Government identity database → biometric verification interface → airport access → passenger-processing market

If competitors cannot reproduce the relevant infrastructure because access depends on a government-controlled or dominant private system, competition authorities may examine whether discriminatory access amounts to exclusion.

However, not every important infrastructure is an essential facility. Security, privacy, national-security and technical-integrity considerations can constitute legitimate reasons for restricting access.

V. Geographic Partitioning

Geographic partitioning is particularly important in cross-border digital markets.

An AI border system could theoretically create:

Country A data → Country A algorithm → Country A provider

while preventing:

Country B provider → Country A customers

This can transform a potentially international digital market into a collection of protected national markets.

Competition concerns become stronger where the geographic separation is created by private commercial conduct rather than unavoidable legal requirements.

VI. Relevant Case Laws

The following cases provide established competition-law principles that can be applied to AI border-control and digital market-partitioning scenarios.

1. United Brands Company v Commission

Case 27/76, Court of Justice of the European Union

The Court examined discriminatory commercial practices and the abuse of a dominant position.

Principle

A dominant undertaking cannot use its market power to impose discriminatory conditions that distort competitive relationships.

Relevance

An AI border-control provider with dominance in biometric infrastructure could potentially raise competition concerns if it gives materially different technical access to competing providers without objective justification.

2. Commercial Solvents Corp v Commission

Joined Cases 6/73 and 7/73

This case concerned a dominant undertaking's refusal to supply an input to a downstream competitor.

Principle

A dominant undertaking controlling an important upstream input cannot necessarily use that control to eliminate downstream competition.

Application

Suppose a dominant AI border platform controls a critical identity-verification component and refuses access to rival border-processing companies.

The case provides a foundation for examining whether such conduct constitutes exclusionary abuse.

3. Oscar Bronner GmbH & Co KG v Mediaprint

Case C-7/97

This is one of the principal European cases concerning refusal of access to infrastructure.

Principle

A refusal to provide access to infrastructure becomes particularly significant where the facility is indispensable and duplication is not realistically possible.

AI-border application

Consider an airport AI verification system that:

• controls access to biometric authentication;
• cannot realistically be duplicated;
• is necessary for competing digital travel providers;
• excludes those competitors from the market.

Bronner provides an important framework for evaluating the indispensability question.

4. IMS Health GmbH & Co OHG v Commission

Joined Cases C-418/01 P and C-7/01 P

The case concerned access to a dominant firm's intellectual-property-protected infrastructure.

Principle

Exceptional circumstances can justify intervention where refusal of access prevents the emergence of a new product or service and eliminates competition.

Relevance

An AI border-control platform may incorporate:

• proprietary APIs;
• biometric databases;
• interoperability protocols;
• software interfaces.

If competitors cannot develop competing services without access to those interfaces, the IMS Health framework becomes relevant.

5. Microsoft Corp v Commission

Case T-201/04

The case concerned interoperability information and Microsoft's position in software markets.

Principle

Control over interoperability information can provide a powerful competitive advantage and can become an instrument of exclusion.

AI border-control application

This is particularly relevant to:

• biometric interoperability;
• API access;
• identity-verification protocols;
• airport software;
• government-cloud integration.

A dominant supplier that intentionally prevents rival systems from interoperating may create a technological partition between its own ecosystem and competing ecosystems.

6. Google Shopping

Case T-612/17, Google and Alphabet v Commission

The European Commission's decision, subsequently reviewed by the General Court, concerned Google's treatment of competing comparison-shopping services.

Principle

A dominant digital platform's design and ranking practices can raise abuse-of-dominance concerns where they disadvantage competing services.

AI border-control application

The analogous concern could arise where an AI border platform:

• ranks its own verification service first;
• systematically demotes competing identity providers;
• gives its affiliated services privileged API access;
• directs users toward its own ecosystem.

The relevant issue is not simply algorithmic decision-making but whether algorithmic architecture is being used to leverage dominance into adjacent markets.

7. Slovak Telekom v Commission

Joined Cases C-152/19 P and C-165/19 P

The case concerned access to telecommunications infrastructure and exclusionary conduct by a dominant undertaking.

Principle

A dominant infrastructure operator can face competition-law scrutiny when access conditions prevent competitors from competing effectively.

Relevance

AI border systems increasingly depend on telecommunications and digital infrastructure.

A dominant operator could potentially use:

• discriminatory access conditions;
• technical limitations;
• pricing structures;
• interoperability requirements

to disadvantage competing providers.

8. Deutsche Telekom v Commission

Case C-280/08 P

This case concerned exclusionary pricing practices involving a dominant telecommunications operator.

Principle

Competition law can address conduct by a dominant undertaking that makes effective competition difficult even when the undertaking formally offers access.

Application

An AI-border technology provider might theoretically comply with an access obligation formally while imposing:

• discriminatory fees;
• technical conditions;
• unreasonable integration costs;
• incompatible technical standards.

The analysis would therefore need to examine effective access, rather than merely nominal access.

VII. Digital Market Partitioning Through APIs

APIs are especially important.

Imagine:

Border-Control Platform A

controls:

• identity API;
• biometric API;
• customs API;
• risk-scoring API.

It allows its own affiliated travel platform full access but gives competitors:

• delayed data;
• incomplete fields;
• restricted call volumes;
• expensive access;
• incompatible formats.

The resulting structure may be:

Data → dominant API → affiliated platform → customers

while competing providers face:

Data ✕ competing API → limited market access

This can potentially constitute digital foreclosure.

VIII. Algorithmic Territorial Discrimination

AI can also partition markets automatically.

For example, an algorithm could assign:

The competition-law question is whether these distinctions result from:

1. genuine legal requirements;
2. security requirements;
3. privacy requirements;
4. technical limitations;

or instead:

1. exclusionary commercial strategy.

The presence of geographic differentiation does not itself establish an infringement.

IX. Data Localisation and Competition

Data localisation can have both legitimate and anticompetitive dimensions.

Legitimate reasons

Governments may require localisation because of:

• national security;
• privacy;
• sovereignty;
• cybersecurity;
• evidentiary requirements.

Competition concern

A dominant technology supplier could potentially exploit localisation requirements by arguing:

"Only our locally established system can process the data."

If the provider then prevents foreign or rival suppliers from accessing interoperable services, localisation may indirectly reinforce market power.

The competition analysis must distinguish government-created barriers from private exploitation of those barriers.

X. AI Border Systems and Self-Preferencing

A particularly important hypothetical structure is:

Border infrastructure provider

↓

operates AI identity system

↓

also owns travel/immigration platform

↓

ranks its own services above competitors.

This resembles concerns examined in digital-platform cases such as Google Shopping.

The important questions include:

• Does the provider have dominance?
• Is the platform vertically integrated?
• Does it favour its own downstream service?
• Are rivals disadvantaged?
• Is the preference technically necessary?
• Can consumers realistically switch?

XI. Tying and Bundling

AI border-control suppliers may offer multiple components:

biometric verification + cloud storage + analytics + customs software

A dominant provider could potentially condition access to one product upon purchasing another.

For example:

"Airport operators receiving biometric verification must also purchase our analytics platform."

Such arrangements may raise tying or bundling concerns if the supplier possesses market power and the arrangement forecloses competitors in the tied market.

XII. Interoperability as a Competition Remedy

Competition authorities may consider interoperability-oriented remedies such as:

1. API access

Competitors receive technically equivalent access.

2. Data portability

Customers can transfer relevant information to another provider.

3. Open technical standards

Interfaces are made compatible with rival systems.

4. Non-discrimination

The dominant provider cannot favour its own downstream service.

5. Functional separation

Infrastructure and competitive downstream services may be separated.

6. Transparency

Providers may be required to explain relevant access conditions without revealing sensitive security information.

XIII. Special Problem of Government Procurement

Government procurement creates an additional dimension.

A government might award a large contract for:

• biometric infrastructure;
• AI risk assessment;
• customs automation;
• identity verification.

A supplier may subsequently obtain a significant installed base.

Competition issues can arise if the supplier then uses that installed base to:

• lock in airports;
• prevent interoperability;
• restrict switching;
• exclude alternative suppliers;
• extend its dominance into adjacent markets.

However, the procurement decision itself may be governed primarily by public procurement and administrative law, rather than competition law.

XIV. Public Authority Versus Private Undertaking

This distinction is fundamental.

Competition law generally focuses on conduct undertaken by an economic undertaking.

A government exercising sovereign border-control functions is not automatically treated like a private technology company.

Therefore, analysis should separate:

Sovereign activity

• immigration control;
• passport inspection;
• national-security screening;
• customs enforcement.

Economic activity

• commercial biometric services;
• private airport technology;
• cloud services;
• software licensing;
• identity-verification platforms;
• commercial data processing.

The same technical system may contain both dimensions.

XV. Security Justifications

AI border-control systems present unusually strong legitimate justifications for restricted access.

Possible justifications include:

• prevention of identity fraud;
• terrorism prevention;
• cybersecurity;
• protection of biometric information;
• prevention of system manipulation;
• integrity of government databases;
• national-security requirements.

Accordingly, competition law should not automatically equate restricted access with unlawful exclusion.

The critical inquiry is proportionality:

Is the restriction genuinely necessary for security, or is security being used as a pretext for commercial foreclosure?

That question must be supported by evidence rather than assumption.

XVI. Cross-Border Digital Market Partitioning

The problem becomes particularly complex when several jurisdictions impose different requirements.

For example:

EU border system

→ GDPR requirements

US border system

→ different data-access requirements

Asian border system

→ localisation requirements

Private AI supplier

→ proprietary interoperability architecture

The resulting market may fragment into:

EU digital identity market
US digital identity market
Asian digital identity market

A dominant provider could then establish separate technological ecosystems and make cross-border competition difficult.

XVII. Competition-Law Test

A structured analysis can be conducted in eight stages:

Step 1 — Define the relevant market

Possible markets include:

• biometric verification;
• AI border-screening software;
• airport identity systems;
• customs AI;
• government identity infrastructure;
• digital travel verification.

Step 2 — Determine market power

Consider:

• market share;
• switching costs;
• network effects;
• data advantages;
• regulatory barriers;
• installed base;
• interoperability.

Step 3 — Identify the partitioning mechanism

Determine whether partitioning results from:

• contract;
• API restriction;
• algorithm;
• data localisation;
• technical incompatibility;
• exclusive dealing.

Step 4 — Establish competitive harm

Examine:

• foreclosure;
• reduced entry;
• higher switching costs;
• reduced innovation;
• increased prices;
• reduced choice.

Step 5 — Examine legitimate justification

Consider:

• security;
• privacy;
• cybersecurity;
• regulatory compliance;
• technical integrity.

Step 6 — Examine proportionality

Could the legitimate objective be achieved through a less restrictive arrangement?

Step 7 — Determine causal connection

Was the market partition caused by:

• law;
• government action;
• technical necessity;
• or private commercial conduct?

Step 8 — Consider remedies

Potential remedies include:

• interoperability;
• API access;
• non-discrimination;
• data portability;
• prohibition of exclusivity;
• structural separation.

XVIII. Relationship With Other Areas of Law

AI border-control systems sit at the intersection of several legal regimes:

Consequently, a competition-law analysis should not treat border-control technology as an ordinary commercial platform.

XIX. Key Doctrinal Lessons From the Cases

The cases collectively demonstrate several important propositions:

1. Control over infrastructure can create competitive leverage.
2. Interoperability can be competitively significant.
3. Refusal of access requires careful indispensability analysis.
4. Dominant firms cannot necessarily use upstream control to eliminate downstream competition.
5. Algorithmic design can produce exclusionary effects.
6. Formal availability of access does not necessarily mean effective competitive access.
7. Geographic differentiation is not automatically unlawful.
8. Legitimate regulatory and security objectives must be distinguished from private commercial foreclosure.

XX. Hypothetical Example

Assume BorderAI Ltd. operates AI biometric verification at 70% of a country's international airports.

It also owns TravelPass, a private digital travel platform.

BorderAI provides:

• full biometric API access to TravelPass;
• limited API access to competing travel platforms;
• faster authentication for TravelPass;
• lower integration costs for its affiliate;
• different technical standards in different countries.

The potential competition issues include:

Dominance

Does BorderAI possess substantial market power?

Leveraging

Is it using border infrastructure dominance to strengthen TravelPass?

Discrimination

Are competitors receiving objectively equivalent access?

Interoperability

Can rival platforms technically connect?

Geographic partitioning

Are national versions deliberately made incompatible?

Tying

Must airports purchase additional BorderAI services to obtain biometric functionality?

Objective justification

Are differences genuinely necessary for security?

The relevant case-law principles would include Bronner, IMS Health, Microsoft, Google Shopping, Slovak Telekom, Deutsche Telekom, United Brands and Commercial Solvents.

XXI. Conclusion

AI border-control systems represent a distinctive form of digitally embedded public infrastructure. Their competition implications arise not merely because AI is used, but because control over identity systems, biometric interfaces, APIs, data and interoperability can create substantial economic power.

The central legal distinction is between:

governmentally required market separation

and

privately engineered market foreclosure.

Competition law may become relevant where a dominant undertaking converts control over border-related digital infrastructure into an advantage in adjacent commercial markets through discriminatory access, refusal to interoperate, tying, self-preferencing, exclusionary contracts or technological incompatibility.