Competition Law And Intelligent Legal Infrastructure Market Powe
Competition Law and Intelligent Legal Infrastructure Market Power
1. Introduction
Intelligent legal infrastructure refers to the technological and institutional systems that support legal research, legal information, case management, e-discovery, compliance, legal analytics, AI-assisted legal research, digital courts, electronic filing, legal databases, automated contract systems and other data-intensive legal services.
The competition-law significance of these systems arises because a platform may control legal data, case-law databases, proprietary taxonomies, citation systems, interoperability interfaces, AI training data, professional workflows, or access to institutional users. Such control can create market power even where the underlying primary law—statutes, regulations and judgments—is publicly available.
The classic example is the U.S. Thomson–West litigation. The DOJ identified competition concerns in enhanced primary-law products, secondary legal publications and comprehensive online legal research. The proposed acquisition raised concerns that reduced competition could lead to higher prices, lower quality and reduced innovation, while access to certain Thomson products was important to LexisNexis's ability to compete.
Modern AI makes the issue more significant because legal-information platforms can combine proprietary content + metadata + editorial classification + search algorithms + AI models + user workflows into an integrated ecosystem.
2. Meaning of Market Power in Intelligent Legal Infrastructure
Market power is the ability of an undertaking to behave to a significant extent independently of competitive constraints or to influence competitors, customers or market conditions.
In intelligent legal infrastructure, market power can arise from:
- Large proprietary legal databases
- Unique case-law metadata
- Citation and classification systems
- AI training datasets
- Network effects
- High switching costs
- Professional dependency
- Integration with law-firm workflows
- Interoperability control
- Control over APIs
- Data accumulation
- Reputation and accuracy advantages
- Regulatory or institutional relationships
- Economies of scale in AI
- Vertical integration between legal content and AI tools
A legal database may therefore become a competitive bottleneck even though individual court judgments themselves remain publicly accessible.
3. Relevant Markets
Competition authorities should avoid defining the market merely as "legal technology." Several narrower markets may exist.
A. Online legal research
Examples include:
- case-law databases;
- statutory databases;
- regulatory databases;
- legal commentary;
- citation services;
- AI-assisted legal research.
B. Legal analytics
This may involve:
- judge analytics;
- litigation outcome data;
- damages analytics;
- law-firm analytics;
- case prediction;
- litigation strategy tools.
C. AI legal research
This emerging market may involve:
- natural-language legal search;
- AI-generated legal research;
- case summarisation;
- legal reasoning tools;
- AI-powered precedent identification.
D. Legal workflow infrastructure
This includes:
- e-filing;
- case management;
- document management;
- contract lifecycle management;
- e-discovery;
- compliance systems.
E. Institutional legal infrastructure
A separate market may arise around technology supplied to:
- courts;
- government departments;
- regulators;
- arbitral institutions;
- prosecutors;
- public legal-information systems.
The appropriate market definition depends on substitutability, functionality, users, geography and competitive constraints rather than the technological label.
4. Sources of Market Power
4.1 Proprietary Legal Data
A platform possessing a large historical collection of:
- judgments;
- annotations;
- headnotes;
- legislative histories;
- citation relationships;
- regulatory materials;
can develop advantages that competitors cannot easily replicate.
The important distinction is between public legal information and value-added proprietary organization of that information.
Westlaw, for example, combines legal materials with editorial features such as headnotes and its Key Number System. The litigation involving Ross Intelligence illustrates how these value-added components can become strategically important to competing AI legal-search products.
4.2 Data Network Effects
An intelligent legal platform can become more useful as it accumulates:
users → queries → data → improved algorithms → better results → more users.
This can generate a feedback loop.
A large incumbent may therefore possess an advantage not simply because it has a large database, but because its historical data and user interactions improve its AI systems.
Competition authorities may need to examine whether such advantages are:
- legitimately earned;
- replicable by rivals;
- protected by technical barriers;
- reinforced by contractual restrictions.
5. AI and Legal Infrastructure
AI creates a new dimension of competition.
A legal AI platform may require:
legal corpus → metadata → training data → model → retrieval system → citation verification → professional workflow.
Control over any one of these components can affect competition in downstream markets.
For example, an incumbent legal-information provider could theoretically possess:
- proprietary case annotations;
- historical legal research data;
- citation networks;
- editorial classifications;
- AI training material;
- legal-document datasets.
A rival attempting to construct a competing system may therefore face significant entry barriers.
The recent Thomson Reuters v. Ross Intelligence litigation demonstrates the importance of this issue. Ross sought to build an AI legal research product, while Thomson Reuters alleged that Westlaw headnotes were improperly used through third-party materials in training the competing system. The 2025 Delaware proceedings resulted in summary judgment findings concerning copyright infringement and fair use; importantly, this was a copyright dispute rather than an antitrust judgment establishing dominance.
6. Essential Facilities and Access
An intelligent legal infrastructure provider may control an input that competitors require.
Potential examples include:
- unique legal databases;
- interoperability protocols;
- court-data interfaces;
- APIs;
- citation databases;
- legally significant metadata;
- institutional authentication systems.
Competition law traditionally treats compulsory access to infrastructure cautiously.
The European jurisprudence provides important principles through the Magill, IMS Health, Bronner and Microsoft line of cases. In Microsoft, the Court accepted that refusal to provide interoperability information could constitute abusive conduct where the relevant conditions were satisfied. The jurisprudence subsequently became important for analysing access to technologically necessary inputs.
7. Six Major Case Laws
7.1 United States v. Thomson Corp. / West Publishing Co. — 1996
Jurisdiction: United States
Law: Section 7, Clayton Act
This is the most directly relevant competition case for legal-information infrastructure.
The DOJ challenged Thomson's proposed acquisition of West Publishing. The government identified numerous relevant product markets, including:
- enhanced primary legal materials;
- secondary legal publications;
- comprehensive online legal research.
The government was concerned that the merger would eliminate competition between important legal-information suppliers. It also considered LexisNexis's dependence on certain Thomson products for effective competition in online legal research.
Principle
Legal-information databases and enhanced legal research products can constitute distinct antitrust markets, and control over particular legal-information inputs can affect downstream competition.
Relevance to intelligent infrastructure
The case provides a direct foundation for analysing:
- AI legal databases;
- legal research platforms;
- proprietary legal metadata;
- citation services;
- database mergers;
- access restrictions.
7.2 Microsoft Corp. v. Commission — EU
Case: Microsoft v Commission, Case T-201/04
Microsoft concerned interoperability information and Microsoft's position in operating systems.
The Court upheld the finding that Microsoft's refusal to provide interoperability information could constitute abusive conduct under Article 82 EC, subject to the applicable legal conditions. The case is particularly significant because interoperability information can constitute a technologically important input for downstream competitors.
Principle
Dominance over an upstream technological infrastructure may create competition concerns where refusal to provide necessary interoperability information prevents effective downstream competition.
Application
Comparable questions could arise if an intelligent legal platform controls:
- essential APIs;
- interoperability protocols;
- court-data interfaces;
- authentication systems;
- machine-readable legal datasets.
7.3 IMS Health v. Commission
Case: IMS Health GmbH & Co. OHG v Commission
IMS Health concerned access to a data structure protected by intellectual-property rights.
The case is important because it addressed the difficult intersection between:
IP rights + market dominance + refusal to license + downstream competition.
The jurisprudence established stringent conditions for treating refusal to license an intellectual-property right as abusive. Later Microsoft jurisprudence developed the analysis concerning technical progress and effective competition.
Relevance
An AI legal platform might own:
- proprietary taxonomies;
- structured datasets;
- annotated legal content;
- proprietary APIs.
Competition law must distinguish legitimate IP protection from circumstances in which control over IP becomes an instrument for excluding effective competition.
7.4 Bronner v Mediaprint
Case: Oscar Bronner GmbH & Co. KG v Mediaprint, Case C-7/97
Bronner is a foundational EU case on refusal to deal and essential facilities.
The Court applied stringent requirements before requiring a dominant firm to provide access to infrastructure.
Principle
Not every commercially important facility is an "essential facility."
A refusal to provide access becomes particularly problematic only when the legal conditions for intervention are satisfied.
Intelligent legal infrastructure
The same reasoning could apply to:
- proprietary legal-data repositories;
- AI interfaces;
- legal-document exchanges;
- court technology;
- specialised legal APIs.
A platform's usefulness alone would not automatically justify mandatory access.
7.5 Magill
Cases: RTE and ITP v Commission, Joined Cases C-241/91 P and C-242/91 P
Magill concerned copyright and refusal to license information.
The case established the classic exceptional circumstances under which refusal to license intellectual property may amount to abuse of dominance.
Principle
Competition law can intervene at the intersection of:
IP rights + dominance + refusal to license
when the demanding conditions identified by the Court are satisfied.
Application to legal AI
The case becomes relevant where an incumbent argues:
"Our legal-information database and AI-related content are protected intellectual property, therefore competitors cannot obtain access."
Competition law does not automatically invalidate that argument, but neither does IP protection automatically immunise exclusionary conduct.
7.6 Matrimony.com Ltd. v. Google LLC & Others — CCI
Jurisdiction: India
Authority: Competition Commission of India
Case Nos.: 07 & 30 of 2012
This case is important for understanding digital infrastructure and dominance under Section 4 of the Competition Act, 2002.
The CCI examined:
- online general web search;
- online search advertising;
- Google's market position;
- search bias;
- access restrictions;
- leveraging between related markets.
The CCI found Google dominant in the relevant online search and search-advertising markets and examined whether certain practices could exploit or reinforce that position.
Principle
Digital market power cannot be assessed solely by looking at conventional price-based indicators. Relevant considerations include:
- network effects;
- technological advantages;
- dependence;
- entry barriers;
- vertical integration;
- market structure;
- access to users.
These considerations are particularly relevant to intelligent legal infrastructure.
8. Additional Relevant Case: Google Android
Google Android — CCI, Case No. 39 of 2018
The CCI's Android proceedings examined Google's position in markets involving:
- licensable mobile operating systems;
- app stores;
- related digital ecosystems.
The CCI treated Google's ecosystem position as relevant to assessing dominance and conduct. Later CCI proceedings continued to rely on the market definitions and dominance findings developed in the Android and Google Play matters.
Relevance
The case demonstrates how competition analysis can move beyond an individual product toward an ecosystem model.
An intelligent legal infrastructure provider could similarly combine:
legal database + AI assistant + document management + analytics + workflow + institutional access.
Competition concerns may arise when an undertaking uses dominance in one layer to strengthen its position in another.
9. Self-Preferencing
An intelligent legal platform may simultaneously operate:
- the underlying legal database;
- the search engine;
- an AI assistant;
- a legal analytics product;
- a marketplace for third-party legal services.
This creates a potential conflict.
For example, an AI legal-search platform could theoretically rank its own:
- commentary;
- AI-generated answers;
- legal products;
- document tools;
- affiliated services
more prominently than competing products.
The competition question would be whether such conduct constitutes exclusionary self-preferencing or another recognised abuse, depending on the applicable jurisdiction and evidence.
10. Tying and Bundling
An incumbent might bundle:
legal database + AI assistant + document management + citation verification + analytics
into one mandatory subscription.
Bundling is not automatically unlawful.
The competition analysis would examine factors such as:
- dominance in the tying product;
- distinctness of products;
- coercion;
- foreclosure;
- duration;
- competitive effects;
- efficiencies;
- availability of alternatives.
This is especially important because AI legal infrastructure is increasingly becoming an integrated ecosystem rather than a standalone database.
11. Exclusive Contracts
An intelligent legal infrastructure provider might enter into agreements with:
- courts;
- government agencies;
- universities;
- major law firms;
- legal publishers;
- arbitral institutions.
Long-term exclusive arrangements can potentially create entry barriers where competitors require access to the same customers or data.
However, exclusivity is not inherently anticompetitive. Its legality depends upon the applicable statute and evidence of foreclosure or other competitive harm.
12. Data Portability and Switching Costs
Legal professionals can accumulate years of:
- saved authorities;
- research folders;
- annotations;
- document histories;
- citation libraries;
- litigation analytics;
- internal precedents.
If these cannot easily be transferred to another provider, users may become locked into a platform.
This creates switching costs.
Competition authorities may therefore consider:
- data portability;
- API access;
- export formats;
- interoperability;
- migration costs;
- contractual restrictions.
13. AI Training Data as a Competitive Bottleneck
AI creates an additional potential source of market power.
Consider:
Court decisions → annotations → training data → legal AI → users → additional data
An incumbent controlling proprietary legal annotations may possess an advantage in developing competing AI systems.
But an important legal distinction must be maintained:
Copyright ownership is not equivalent to antitrust dominance.
Similarly:
Possession of valuable data is not automatically an essential facility.
Competition law requires analysis of the relevant market and the actual competitive effects.
The Ross litigation illustrates why proprietary legal content can be strategically important to competing AI systems, although the case itself was a copyright dispute rather than an antitrust determination.
14. Predatory or Exclusionary Pricing
An established legal-information platform could potentially price an AI product aggressively to prevent entry.
Competition analysis would examine:
- cost structure;
- pricing duration;
- recoupment where legally relevant;
- exclusionary intent/effect;
- competitive alternatives;
- efficiencies.
The fact that an AI legal product is initially free or inexpensive does not, by itself, establish predatory pricing.
15. Algorithmic Competition
Intelligent legal platforms increasingly use algorithms to determine:
- search rankings;
- relevant precedents;
- legal authorities;
- document recommendations;
- litigation analytics;
- contract-risk scores.
Algorithms can therefore influence competitive conditions.
Potential concerns include:
A. Algorithmic self-preferencing
The system systematically prioritises affiliated products.
B. Algorithmic exclusion
Competitors' content is systematically disadvantaged.
C. Coordinated algorithms
Separate platforms could theoretically use algorithms in ways that facilitate coordination.
D. Information asymmetry
The platform may possess significantly more information about competitors and customers than users possess about the platform.
16. Institutional Dependence
Intelligent legal infrastructure has a special characteristic: users may not be ordinary consumers.
Customers can include:
- courts;
- regulators;
- government agencies;
- major law firms;
- corporations;
- arbitral institutions.
A platform that becomes embedded in institutional workflows can acquire significant dependency-based market power.
For example:
Court system → API → case database → legal research platform → law firms → litigation workflow
If the platform becomes indispensable to the ecosystem, competition authorities may need to examine interoperability and access conditions.
17. Merger Control
Mergers in intelligent legal infrastructure deserve particular scrutiny where parties combine:
- legal databases;
- AI models;
- legal analytics;
- e-discovery;
- contract platforms;
- court-data services.
Traditional concentration measures may not fully capture the competitive significance of:
- data;
- innovation;
- interoperability;
- nascent competition;
- potential entrants.
The Thomson–West litigation remains particularly instructive because the government identified numerous distinct legal-information markets rather than treating all legal publishing as one market.
18. Potential Theories of Harm
| Conduct | Possible competition concern |
|---|---|
| Acquisition of rival legal database | Elimination of competition |
| Refusal to license proprietary data | Foreclosure |
| API denial | Interoperability restriction |
| Self-preferencing | Competitor exclusion |
| Bundling AI + database | Leveraging |
| Exclusive court contracts | Entry barriers |
| Predatory AI pricing | Exclusion of entrants |
| Data portability restrictions | Customer lock-in |
| Exclusive training-data agreements | Input foreclosure |
| Algorithmic ranking manipulation | Discriminatory access |
| Vertical integration | Raising rivals' costs |
| Acquisition of nascent AI competitor | Elimination of future competition |
19. Remedies
Where competition concerns are established, possible remedies may include:
Structural remedies
- divestiture;
- separation of businesses;
- prohibition of particular acquisitions.
Behavioural remedies
- non-discriminatory access;
- interoperability obligations;
- licensing;
- API access;
- data portability;
- non-exclusivity.
Technical remedies
- open standards;
- interoperable formats;
- export functionality;
- transparent ranking systems;
- audit mechanisms.
AI-specific remedies
- restrictions on discriminatory model access;
- training-data access under appropriate conditions;
- safeguards against self-preferencing;
- independent auditing;
- model-output transparency where necessary.
Remedies should be proportionate to the competition problem actually established.
20. Indian Legal Framework
The principal statutory provisions are:
Section 3
Prohibits anti-competitive agreements.
Relevant possibilities include:
- exclusive arrangements;
- information-sharing;
- coordinated conduct;
- restrictive licensing arrangements.
Section 4
Deals with abuse of dominant position.
Potentially relevant forms include:
- unfair conditions;
- denial of market access;
- discriminatory treatment;
- leveraging;
- tying/bundling.
Sections 5 and 6
Govern combinations and merger control.
Section 19
Provides the framework for CCI investigations.
Section 19(4) expressly permits consideration of factors including:
- market share;
- size and resources;
- economic power;
- vertical integration;
- consumer dependence;
- entry barriers;
- technical barriers;
- economies of scale;
- market structure.
These factors are particularly relevant to AI-based legal infrastructure. The CCI relied on such considerations in analysing Google's dominance in online search and search advertising.
21. Competition Assessment Framework
A regulator examining intelligent legal infrastructure can proceed through the following sequence:
Step 1 — Identify the technology
↓
Step 2 — Define the relevant product/service market
↓
Step 3 — Define the geographic market
↓
Step 4 — Measure market power
↓
Step 5 — Examine data and network effects
↓
Step 6 — Identify essential inputs
↓
Step 7 — Examine interoperability
↓
Step 8 — Analyse exclusionary conduct
↓
Step 9 — Examine effects on rivals and consumers
↓
Step 10 — Consider efficiencies and innovation
↓
Step 11 — Determine proportionate remedy
This prevents the analysis from equating technological sophistication with dominance.
22. Key Competition-Law Questions
For intelligent legal infrastructure, the central questions are:
- What is the relevant market?
- Who controls the critical legal data?
- Can competitors realistically reproduce the database?
- Are legal datasets substitutable?
- Are proprietary annotations replicable?
- Does the platform benefit from network effects?
- Are customers locked in?
- Can data be ported?
- Are APIs available on reasonable terms?
- Does the provider favour its own downstream products?
- Does bundling foreclose rivals?
- Does an acquisition eliminate an emerging competitor?
- Does the platform control a genuine essential input?
- Are restrictions objectively justified?
- Would intervention improve competition without undermining legitimate innovation?
23. Important Distinction: Legal Information vs Legal Infrastructure
A crucial distinction should be made between:
Public legal information
- judgments;
- statutes;
- regulations;
- publicly available government materials.
and
Intelligent legal infrastructure
- proprietary indexing;
- annotations;
- headnotes;
- citation analytics;
- AI retrieval;
- predictive analytics;
- document intelligence;
- proprietary APIs;
- workflow integration.
Competition concerns are generally stronger where a firm controls the infrastructure surrounding the information, rather than merely possessing information that is freely available from public sources.
24. Overall Legal Position
The competition-law problem presented by intelligent legal infrastructure is therefore not simply:
"Does a company have a large legal database?"
The more important question is:
"Does control over legal data, intelligence, interoperability, AI capabilities or institutional workflows allow the undertaking to exercise market power or exclude effective competition?"
The Thomson–West matter provides the most direct precedent because it involved competition in legal research and legal publishing itself. Magill, IMS Health, Bronner and Microsoft provide the broader EU framework for analysing access, interoperability, intellectual property and essential-facility-type issues. Matrimony.com v Google and the CCI's Google ecosystem cases demonstrate how similar principles can operate in India's digital-platform environment.
The emerging AI legal infrastructure therefore combines traditional competition-law concerns—dominance, foreclosure, tying, exclusivity and refusal to deal—with newer issues involving training data, algorithmic ranking, interoperability, data portability, network effects and ecosystem dependency.
Key Case Laws at a Glance
- United States v. Thomson Corp. / West Publishing Co., 949 F. Supp. 907 (D.D.C. 1996) — legal research and publishing merger.
- Microsoft Corp. v. Commission, Case T-201/04 — interoperability and refusal to supply technological information.
- IMS Health GmbH & Co. OHG v. Commission, Joined Cases C-418/01 P & related proceedings — IP, access and dominance.
- Oscar Bronner GmbH & Co. KG v. Mediaprint, Case C-7/97 — essential facilities/refusal to deal.
- RTE & ITP v. Commission (Magill), Joined Cases C-241/91 P & C-242/91 P — IP licensing and abuse of dominance.
- Matrimony.com Ltd. v. Google LLC & Others, CCI Case Nos. 07 & 30 of 2012 — digital search dominance and leveraging.
- Google Android, CCI Case No. 39 of 2018 — ecosystem dominance and digital-platform leveraging.
- Thomson Reuters Enterprise Centre GmbH v. Ross Intelligence Inc. — not an antitrust case, but highly relevant to the emerging relationship between proprietary legal information, AI training and competitive legal-research infrastructure.
Exam takeaway: Intelligent legal infrastructure can acquire market power through control of proprietary legal data, interoperability, AI capabilities, network effects and institutional dependency. Competition law must therefore assess not only market share but also access, switching costs, data advantages, ecosystem effects and the possibility that control of one layer of legal technology can be leveraged into another.

comments