Competition Law and Future Regulation of Knowledge-Based Dominance

Introduction

Knowledge-based dominance refers to market power arising primarily from control over valuable knowledge assets rather than traditional physical infrastructure or financial capital. Such knowledge may include proprietary databases, algorithms, AI models, trade secrets, technical standards, user-generated information, business intelligence, software architectures, datasets, scientific know-how, and accumulated learning or expertise.

Modern digital and knowledge-intensive markets create a distinctive competition problem: a firm may become powerful because it possesses a large and continuously improving knowledge advantage, and that advantage can become self-reinforcing. More users generate more data; more data improves the product; the improved product attracts more users; and the resulting knowledge advantage makes entry increasingly difficult.

Competition law therefore faces the question of how to distinguish:

• legitimate rewards for innovation and investment;
• intellectual-property protection;
• legitimate confidentiality and trade-secret protection; from
• strategic withholding of knowledge that substantially forecloses competition.

The future regulatory challenge will be particularly significant in AI, cloud computing, digital platforms, pharmaceuticals, biotechnology, financial technology, professional services, industrial software and data-driven markets.

I. Meaning of Knowledge-Based Dominance

Knowledge-based dominance occurs where a firm's competitive strength substantially depends upon its possession or control of knowledge that rivals cannot readily reproduce.

Important sources include:

1. Proprietary datasets
2. Algorithms and machine-learning models
3. Trade secrets
4. Technical know-how
5. Customer and transaction data
6. Software source code and system architecture
7. Patents and accumulated intellectual property
8. Industry-specific expertise
9. Standards-related technical information
10. Knowledge generated through network effects

A useful conceptual distinction is:

Knowledge advantage becomes a competition concern when it is transformed into a durable barrier to effective competition.

Possession of knowledge by itself does not constitute an antitrust violation.

II. Why Knowledge Can Produce Market Power

1. High replication costs

A competitor may technically be able to reproduce a product but may lack the years of accumulated information necessary to reach comparable quality.

For example, an AI company may possess:

• years of training data;
• model weights;
• proprietary evaluation data;
• user feedback;
• engineering knowledge;
• behavioural datasets.

The rival therefore faces a much higher entry cost.

2. Data feedback loops

Digital markets can generate a particularly powerful cycle:

Users → Data → Better product → More users → More data

This is sometimes described as a data-network effect.

The competition concern arises where the incumbent's informational advantage becomes sufficiently large that competitors cannot obtain comparable inputs.

3. Learning effects

A firm may improve through continuous interaction with customers.

For example:

1 million transactions → better prediction → better service → more customers → 10 million transactions → further improvement.

The resulting advantage is not simply the possession of a static database. It is accumulated organizational knowledge.

4. Switching costs

Knowledge can also become embedded in a customer's systems.

Examples include:

• proprietary software configurations;
• historical business data;
• customised AI models;
• workflow knowledge;
• platform-specific reputation;
• proprietary APIs.

Customers may consequently remain with an incumbent even when competing products exist.

III. Competition Law Theories Applicable to Knowledge-Based Dominance

Knowledge-based dominance can implicate several established competition-law doctrines.

A. Abuse of dominance

A dominant undertaking may potentially abuse its position through:

• refusal to provide indispensable information;
• discriminatory access to data;
• tying;
• bundling;
• exclusive dealing;
• self-preferencing;
• interoperability restrictions;
• discriminatory licensing;
• exploitative data practices where legally relevant;
• exclusionary contractual restrictions.

B. Essential facilities

In exceptional circumstances, a proprietary knowledge resource may become sufficiently indispensable to competition that refusal of access raises an essential-facilities issue.

However, competition law traditionally applies this doctrine cautiously because compulsory access can weaken incentives to invest.

C. Intellectual-property abuse

IP rights create legitimate exclusivity.

But competition law may intervene where IP rights are used as instruments of exclusion beyond the legitimate competitive scope of the right.

The central tension is:

Innovation incentive vs. competitive access.

D. Refusal to license

A refusal to license proprietary technology or information is generally not automatically unlawful.

The strongest competition concerns arise where factors such as the following are established:

• indispensability;
• elimination of effective competition;
• lack of objective justification;
• consumer harm;
• exceptional circumstances surrounding the refusal.

IV. Major Case Laws

1. Magill TV Guide/Commission v. ITP, BBC and RTÉ

Court: Court of Justice of the European Communities
Year: 1995

Facts

Television broadcasters possessed copyright over programme information. Magill sought to publish comprehensive weekly television listings.

The broadcasters refused to license the information.

Decision

The Court recognised that refusal to license intellectual property could, in exceptional circumstances, constitute an abuse of dominance.

The case established important criteria concerning:

• indispensability;
• prevention of a new product;
• absence of justification;
• elimination of competition.

Importance for knowledge-based dominance

Magill established an important foundation for future disputes involving proprietary information as a competitive resource.

The principle is particularly relevant to:

• proprietary databases;
• technical information;
• API access;
• AI training information;
• industry datasets.

2. IMS Health GmbH & Co. OHG v. NDC Health GmbH

Court: Court of Justice of the European Union
Year: 2004

Facts

IMS Health operated a system for pharmaceutical sales information. Its structure became an important industry standard for analysing pharmaceutical sales.

A competitor sought access to the relevant structure.

Decision

The Court reaffirmed the exceptional circumstances required before a refusal to license intellectual property may constitute abuse.

The Court emphasised factors including:

1. indispensability;
2. elimination of competition;
3. prevention of a new product for which there was consumer demand;
4. lack of objective justification.

Importance

IMS Health demonstrates that information architecture itself can acquire competitive significance.

A proprietary classification system, database structure or technical format can potentially become a bottleneck when competitors cannot realistically compete without access.

3. Microsoft Corp. v. Commission

Court: General Court of the European Union
Year: 2007

Facts

Microsoft was found to have abused its dominant position partly through its refusal to provide interoperability information concerning its work-group server operating systems.

Decision

The European courts upheld the Commission's intervention concerning interoperability information.

Importance

The case is highly relevant to modern knowledge-based markets because it demonstrates that technical information necessary for interoperability can become competitively significant.

Its principles are relevant to:

• APIs;
• cloud interoperability;
• operating systems;
• enterprise software;
• AI-agent interoperability;
• digital ecosystems.

4. Bronner v. Mediaprint

Court: Court of Justice of the European Union
Year: 1998

Facts

Mediaprint operated an extensive newspaper home-delivery system. Bronner sought access to that system.

Decision

The Court applied a strict approach to compulsory access and held that the relevant infrastructure was not sufficiently indispensable under the circumstances.

Importance

Bronner provides an important counterbalance to Magill and IMS Health.

It demonstrates that:

Being useful or advantageous to competitors is not enough to justify compulsory access.

This principle will remain important where firms seek access to proprietary databases, AI models or technical knowledge.

5. Slovak Telekom v. European Commission

Court: Court of Justice of the European Union
Year: 2021

Facts

Slovak Telekom was accused of conduct restricting access by alternative operators to infrastructure necessary for competing in telecommunications markets.

Decision

The Court considered the interaction between refusal-of-access principles and Article 102 TFEU.

Importance

The case demonstrates the importance of analysing:

• the nature of access;
• existing regulatory obligations;
• competitive foreclosure;
• infrastructure dependence.

Knowledge-economy relevance

In modern digital markets, the equivalent of physical network access may involve:

• data access;
• interoperability;
• APIs;
• cloud infrastructure;
• technical documentation.

Thus, the underlying competition question can evolve from physical access to informational access.

6. Google Shopping

European Commission decision: 2017
General Court: 2021
Court of Justice: 2024

Facts

The European Commission found that Google had systematically favoured its comparison-shopping service in search results while applying less favourable treatment to competing comparison-shopping services.

Competition concern

The central issue involved the interaction between:

• search infrastructure;
• algorithms;
• visibility;
• traffic;
• data;
• platform power.

Importance for knowledge-based dominance

Google Shopping illustrates how an algorithmically controlled information environment can influence competitive opportunities.

The relevant resource is not merely a physical facility. It is the algorithmic organisation and distribution of knowledge and attention.

This has implications for:

• search;
• recommendation systems;
• AI assistants;
• marketplace ranking;
• digital advertising;
• content discovery.

7. Google Android

European Commission: 2018
General Court: 2022

The Android case involved Google's contractual practices concerning Android devices, including tying and restrictions affecting competing search and browser services.

The case demonstrates how control over a digital ecosystem can allow a dominant undertaking to leverage one knowledge-rich platform into neighbouring markets.

Its broader significance lies in understanding ecosystem power, rather than examining individual products in isolation.

8. Qualcomm

European Commission decision: 2018

The Qualcomm case concerned payments and exclusivity-related arrangements involving baseband chipsets.

Although it is not a pure "knowledge dominance" case, it is useful because sophisticated technological ecosystems often combine:

• patents;
• technical know-how;
• standards;
• licensing;
• network effects;
• contractual exclusivity.

It demonstrates why future regulation may need to consider the combined effect of IP, technical knowledge and contractual market power.

V. Knowledge-Based Dominance in Artificial Intelligence

AI is likely to become one of the most important areas for future competition regulation.

A leading AI firm may simultaneously control:

• training datasets;
• computational infrastructure;
• model architecture;
• model weights;
• evaluation datasets;
• reinforcement-learning feedback;
• developer ecosystems;
• APIs;
• distribution channels.

This creates multiple potential bottlenecks.

Example

Suppose an AI company has:

proprietary training data + dominant cloud infrastructure + leading model + exclusive distribution + developer ecosystem.

A rival may technically be able to build another AI system, but it may not be able to reproduce the same knowledge accumulation.

The competition question therefore becomes:

Should competition law regulate merely the final AI product, or should it examine control over the underlying knowledge inputs?

VI. Future Regulation of Knowledge-Based Dominance

1. Data portability

Competition authorities may increasingly support mechanisms allowing users or businesses to transfer relevant data between competing services.

Portability can reduce:

• switching costs;
• lock-in;
• informational asymmetry.

However, portability must be balanced against:

• privacy;
• cybersecurity;
• trade secrets;
• third-party rights.

2. Interoperability obligations

Future regulation may require dominant digital ecosystems to provide reasonable interoperability.

Potential areas include:

• messaging;
• cloud services;
• enterprise software;
• payment systems;
• AI agents;
• digital identity;
• IoT ecosystems.

The objective is to prevent technical incompatibility from becoming an artificial barrier to entry.

3. API access regulation

APIs can function as gateways to knowledge and functionality.

A dominant platform may control access to:

• user data;
• payment infrastructure;
• maps;
• search;
• identity;
• advertising information;
• platform functionality.

Future regulation may therefore distinguish between:

legitimate API security restrictions and strategic API foreclosure.

4. Algorithmic transparency

Competition authorities are unlikely to require complete disclosure of commercially sensitive algorithms in every situation.

Instead, future regulation may focus on:

• auditability;
• explanation of discriminatory ranking;
• documentation;
• independent testing;
• access for regulators;
• preservation of evidence.

This represents a movement from source-code disclosure toward accountability and auditability.

5. Data-access remedies

Where a dominant firm's data advantage materially prevents competition, authorities could consider remedies such as:

• data-sharing obligations;
• secure data rooms;
• anonymised datasets;
• API access;
• interoperability;
• independent data trustees.

Such remedies would need to protect:

• privacy;
• cybersecurity;
• trade secrets;
• confidential business information.

6. Knowledge portability

A future regulatory concept could be knowledge portability.

This would go beyond ordinary data portability.

It could involve transferring:

• user preferences;
• historical interactions;
• workflow configurations;
• trained personal models;
• business rules;
• reputation information;
• machine-readable settings.

This would reduce the competitive significance of accumulated user-specific knowledge.

7. Restrictions on exclusive knowledge arrangements

Competition authorities may scrutinise agreements that prevent competitors from obtaining critical knowledge.

Examples could include:

• exclusive data-sharing contracts;
• exclusive AI-training arrangements;
• exclusive technical standards;
• long-term information-sharing restrictions;
• exclusive access to critical datasets.

The analysis would depend upon the actual market effects and legitimate business justifications.

8. Knowledge-sharing remedies in mergers

Traditional merger remedies frequently focus on:

• divestiture;
• assets;
• facilities;
• businesses.

Knowledge-intensive mergers may require different remedies.

Potential remedies include:

• licensing;
• data-access commitments;
• interoperability;
• technical documentation;
• firewall requirements;
• API access;
• restrictions on combining sensitive datasets.

VII. Competition Risks From Knowledge Monopolization

Knowledge-based dominance can produce several forms of competitive harm.

1. Entry barriers

Competitors cannot reproduce the incumbent's accumulated knowledge.

2. Innovation foreclosure

Competitors may lack access to information necessary to develop new products.

3. Reduced interoperability

Closed systems prevent competing products from interacting effectively.

4. Data exclusion

The incumbent prevents rivals from obtaining commercially necessary information.

5. Self-preferencing

The platform uses informational advantages to favour its own downstream products.

6. Leveraging

Knowledge accumulated in one market is used to obtain dominance in another.

7. Exploitative dependency

Smaller firms become dependent upon the dominant undertaking for critical information.

VIII. Limits on Regulation

Over-regulation could itself create competition problems.

1. Reduced innovation incentives

If every successful innovation immediately triggers compulsory access, firms may have less incentive to invest in:

• research;
• data collection;
• AI development;
• proprietary technology.

2. Trade-secret protection

Businesses legitimately need confidentiality.

Compulsory disclosure may destroy the commercial value of knowledge.

3. Privacy

Data access cannot ignore:

• personal-data protection;
• consent;
• purpose limitation;
• data minimisation.

4. Cybersecurity

Opening technical systems to competitors may increase security risks.

5. Free-riding

Competitors should not automatically receive the fruits of another firm's investment.

Consequently, the future regulatory model is likely to focus on targeted access rather than general compulsory sharing.

IX. Future Competition-Law Test for Knowledge-Based Dominance

A useful analytical framework can be constructed around eight questions:

Step 1 — Identify the knowledge asset

What exactly is controlled?

• data;
• algorithm;
• model;
• technical information;
• expertise;
• interoperability information;
• IP.

Step 2 — Define the relevant market

Determine whether the knowledge asset affects:

• the primary market;
• an adjacent market;
• an upstream market;
• a downstream market.

Step 3 — Determine market power

Consider:

• market share;
• entry barriers;
• network effects;
• switching costs;
• data advantages;
• technological lead.

Step 4 — Assess indispensability

Can competitors realistically obtain or reproduce the relevant knowledge?

Step 5 — Examine conduct

Possible conduct includes:

• refusal;
• discrimination;
• tying;
• exclusivity;
• self-preferencing;
• interoperability restrictions.

Step 6 — Assess foreclosure

Does the conduct materially restrict competitors' ability to compete?

Step 7 — Examine justification

Possible legitimate explanations include:

• privacy;
• cybersecurity;
• IP protection;
• trade-secret protection;
• technical limitations;
• efficiency.

Step 8 — Select proportionate remedy

Possible remedies include:

Behavioural remedies → interoperability → data portability → licensing → access obligations → structural remedies in exceptional circumstances.

X. Comparative Regulatory Direction

XI. Relationship Between IP Law and Competition Law

Knowledge-based dominance creates an important boundary between intellectual-property law and competition law.

IP law asks:

How should society reward and protect innovation?

Competition law asks:

How should market power derived from innovation be prevented from being used to suppress competition?

Neither system should automatically override the other.

A patent, copyright, trade secret or database right does not automatically establish unlawful dominance.

Conversely, the existence of IP protection does not necessarily immunise exclusionary conduct from competition scrutiny.

XII. Future Role of Competition Authorities

Competition authorities will increasingly need specialist capabilities in:

• data science;
• AI auditing;
• algorithmic economics;
• cybersecurity;
• software architecture;
• technical standards;
• econometrics;
• machine-learning systems.

Traditional market-share analysis may be insufficient.

Authorities may increasingly investigate:

data concentration + learning effects + interoperability + switching costs + ecosystem control + algorithmic advantages.

XIII. Key Principles Emerging From the Case Law

The major cases collectively support several important propositions:

1. Proprietary information can have competition significance.
2. IP rights do not create unlimited immunity from competition law.
3. Compulsory access remains exceptional.
4. Indispensability is important when access is sought.
5. Interoperability can be a significant competition issue.
6. Digital ecosystems can allow market power to be leveraged across markets.
7. Competition analysis must distinguish legitimate innovation from exclusionary conduct.
8. Remedies must be proportionate to the competitive harm.

XIV. Conclusion

The future of competition law will increasingly involve a shift from asset-based dominance to knowledge-based dominance.

The most significant competitive resources of the future may not be factories, warehouses or physical networks. They may be:

• datasets,
• algorithms,
• AI models,
• technical standards,
• proprietary expertise,
• behavioural information,
• interoperability protocols,
• accumulated learning.

The jurisprudence beginning with Magill, IMS Health and Bronner, followed by Microsoft, Google Shopping, Google Android and other digital-platform cases, provides important foundations for addressing these developments.

However, future regulation must avoid treating every proprietary knowledge advantage as an antitrust problem. The central legal task will be to identify circumstances in which control over knowledge creates durable market power and is strategically used to prevent effective competition.

The likely regulatory direction is therefore toward a combination of competition law, interoperability rules, data portability, targeted access remedies, algorithmic accountability, merger scrutiny and sector-specific digital regulation, while preserving legitimate incentives for innovation, investment, confidentiality and intellectual-property protection.

Key Case Laws at a Glance

1. Magill TV Guide Ltd v ITP, BBC & RTÉ (1995) — exceptional refusal to license intellectual property.
2. IMS Health v NDC Health (2004) — indispensable information and IP licensing.
3. Bronner v Mediaprint (1998) — strict approach to compulsory access.
4. Microsoft v Commission (2007) — interoperability information and exclusionary conduct.
5. Google Shopping (2017/2021/2024) — algorithmic self-preferencing and platform power.
6. Google Android (2018/2022) — ecosystem leverage, tying and contractual restrictions.
7. Slovak Telekom (2021) — access obligations and exclusionary conduct.
8. Qualcomm (2018) — technology ecosystems, exclusivity and market foreclosure.