Competition Law and Genomic Database Concentration Concerns

1. Introduction

Genomic databases are becoming strategically important assets in healthcare, biotechnology, pharmaceutical research, precision medicine, ancestry testing, diagnostics and artificial-intelligence-based drug discovery. Unlike ordinary commercial databases, genomic databases can contain highly valuable and difficult-to-replicate information about genetic sequences, variants, phenotypes, family relationships and health characteristics.

Competition concerns arise when a single undertaking, or a small group of undertakings, obtains control over a very large genomic database through:

• mergers and acquisitions;
• exclusive licensing;
• vertical integration;
• long-term data-sharing agreements;
• acquisition of genetic-testing companies;
• control over sequencing infrastructure;
• aggregation of clinical and genomic datasets;
• interoperability restrictions;
• refusal to provide access to essential datasets; or
• discriminatory access to genomic information.

There is not yet a large body of reported competition cases dealing exclusively with genomic databases. The most directly relevant precedent is Illumina/GRAIL, while cases involving health, personal, user and commercially valuable datasets provide important analogies.

2. Meaning of Genomic Database Concentration

A genomic database concentration occurs where an undertaking obtains or increases control over a significant collection of genomic information.

For example:

Company A operates a genetic-sequencing platform and acquires Company B, which owns a large database containing genomic profiles linked to clinical outcomes.

The transaction may not merely combine two conventional businesses. It may combine:

Genomic data + sequencing technology + diagnostic applications + AI models + clinical databases + distribution channels.

This creates the possibility of data-driven market power.

3. Why Genomic Databases Are Competition-Sensitive

A. High entry barriers

A large genomic database may have taken years to build.

A new entrant may need:

• millions of samples;
• consented participants;
• sequencing infrastructure;
• clinical partnerships;
• computational infrastructure;
• quality-control systems;
• longitudinal health records;
• disease-specific datasets; and
• substantial research expenditure.

Consequently, even where the database itself is technically reproducible, the scale, diversity and historical depth of the data may not be easily replicated.

B. Data network effects

The value of genomic data may increase with the number of individuals and associated clinical outcomes.

For example:

More genomic samples → better statistical models → better diagnostic predictions → more users and clinical partners → more samples → better models.

This creates a feedback loop capable of reinforcing an incumbent's position.

C. Data-driven economies of scope

A company controlling genomic information may use it in multiple neighbouring markets:

• genetic testing;
• pharmaceutical research;
• drug discovery;
• companion diagnostics;
• cancer detection;
• personalised medicine;
• clinical trials;
• AI healthcare applications;
• ancestry services;
• insurance-related analytics, subject to applicable law.

Thus, concentration in one database market can affect competition in several downstream markets.

4. Principal Competition Concerns

4.1 Horizontal concentration

Two competing genomic-data providers may merge.

Potential consequences include:

• reduction in the number of independent datasets;
• elimination of an important competitor;
• increased data concentration;
• reduction in innovation;
• higher prices for researchers;
• reduced licensing options; and
• diminished incentives to improve data quality.

Traditional market-share analysis may therefore be insufficient.

Authorities may also need to examine:

• number of unique genomic records;
• uniqueness of the dataset;
• number of disease categories;
• geographic coverage;
• ethnic/population diversity;
• longitudinal depth;
• linkage to clinical information;
• frequency of updating;
• data quality; and
• replicability.

4.2 Vertical foreclosure

Suppose a sequencing-company acquires a genomic database.

The combined company could potentially:

1. provide preferential access to its own downstream diagnostic business;
2. raise prices for rival researchers;
3. delay access;
4. degrade interoperability;
5. impose restrictive licensing conditions; or
6. deny competitors access to particularly valuable datasets.

This resembles traditional vertical foreclosure, but the input is data rather than a physical product.

4.3 Raising rivals' costs

A dominant genomic-data provider could potentially make competing firms pay more for:

• database access;
• API access;
• data licences;
• sequencing;
• analytical tools;
• clinical datasets; or
• computational services.

Even without completely refusing access, discriminatory pricing or contractual restrictions could make competitors less effective.

4.4 Data exclusivity

Exclusive genomic-data agreements can create significant competitive problems.

For example:

A pharmaceutical company obtains exclusive access to a large disease-specific genomic database for 15 years.

Potential effects include:

• competitors cannot conduct comparable research;
• alternative datasets become less valuable;
• new entrants face higher research costs;
• innovation may become concentrated around the exclusive licensee.

The competition analysis must nevertheless distinguish legitimate incentives to invest in data collection from exclusionary exclusivity.

4.5 Loss of innovation

Genomic markets are strongly innovation-driven.

Competition may occur through:

• better sequencing;
• improved variant interpretation;
• better diagnostic accuracy;
• new biomarkers;
• improved AI models;
• faster drug discovery;
• improved clinical-trial recruitment.

A concentration may therefore be problematic even where immediate consumer prices do not increase.

This is particularly important because genomic services are often supplied at low or zero monetary prices while competition occurs through quality, privacy, research output and innovation.

5. Six Important Case Laws

Case 1: Illumina, Inc. v European Commission / Illumina–GRAIL

Joined Cases C-611/22 P and C-625/22 P, Judgment of 3 September 2024

This is the most important modern precedent for genomic-data-related concentration analysis.

Illumina, a major provider of genetic-sequencing technology, sought to acquire GRAIL, which was developing blood-based multi-cancer early-detection technology. The transaction raised concerns because Illumina supplied sequencing technology relevant to GRAIL and competing future diagnostic products. The FTC separately challenged the transaction in the United States.

The EU litigation ultimately concerned an important jurisdictional issue: whether the European Commission could accept an Article 22 referral from a national competition authority where that authority itself lacked competence to examine the transaction under its national merger-control thresholds. The Court of Justice held that the Commission could not do so in those circumstances and annulled the relevant referral decisions.

Competition significance

The broader Illumina/GRAIL episode demonstrates why small-turnover/high-value biotechnology transactions can present merger-control difficulties.

For genomic databases, the lesson is particularly significant:

• turnover may underestimate competitive significance;
• innovative genomic companies may have limited current revenue;
• control of an important genomic input can have downstream consequences;
• vertical relationships can matter even when the parties are not traditional horizontal competitors;
• innovation competition may be more important than current sales.

The FTC's substantive proceedings found substantial evidence supporting its anticompetitive-concern theory and Illumina ultimately announced that it would divest GRAIL.

Relevance: Extremely high.

Case 2: Google/Fitbit, Case M.9660

The European Commission's Google/Fitbit decision is one of the clearest precedents concerning concentration of sensitive health-related datasets.

Google acquired Fitbit, which possessed substantial health and fitness information. The Commission examined whether combining Google's existing databases with Fitbit's data could strengthen Google's position in online advertising and digital healthcare.

The Commission concluded that Fitbit data could be valuable, particularly in advertising, but found that it was not sufficiently unique to create a significant impediment to competition in the assessed markets. The transaction was conditionally approved, including a data-silo commitment.

Genomic relevance

The case demonstrates that merger authorities can examine:

• the uniqueness of datasets;
• ability to combine datasets;
• data-driven barriers to entry;
• downstream leveraging;
• privacy-related data characteristics; and
• data-silo remedies.

A genomic database may be considerably more difficult to replicate than ordinary fitness data, particularly where it contains genotype–phenotype correlations.

Relevance: Very high.

Case 3: IQVIA/Propel Media

The FTC's challenge to IQVIA's proposed acquisition of Propel Media is particularly relevant to health-data concentration.

IQVIA was described by the FTC as the world's largest healthcare data provider. The agency alleged that the transaction would strengthen IQVIA's position in healthcare programmatic advertising and increase its ability and incentive to leverage important healthcare datasets against rivals.

The FTC specifically identified provider-identity and prescribing-behaviour data as important inputs for competitors.

A federal district court granted a preliminary injunction preventing completion of the acquisition in December 2023. The administrative proceeding was subsequently dismissed in February 2024 after the parties withdrew the matter.

Genomic relevance

The case illustrates the strategic-input theory:

Control over a critical healthcare dataset can create competitive power beyond the market in which the dataset was originally collected.

For genomic databases, the same theory could apply to:

• disease-specific genomic information;
• patient-linked sequencing information;
• pharmacogenomic databases;
• rare-disease datasets; and
• genomic datasets used for AI drug discovery.

Relevance: Very high.

Case 4: Microsoft/LinkedIn, Case M.8124

The European Commission examined Microsoft's acquisition of LinkedIn, including potential effects arising from the combination of Microsoft's ecosystem and LinkedIn's professional data.

The transaction was ultimately declared compatible with the internal market.

Genomic relevance

The case is important because it illustrates how competition authorities may examine data advantages arising from combining datasets held by firms active in adjacent markets.

In genomic markets, similar questions arise where:

• a cloud provider acquires a genomic database;
• an AI company acquires a genetic-testing platform;
• a pharmaceutical company acquires a population-genomics database.

The central question is not simply:

"How large is the database?"

It is:

"What additional competitive capability does the acquiring undertaking obtain by combining this database with its existing assets?"

Relevance: Medium to high.

Case 5: Facebook/WhatsApp, Case M.7217

The European Commission approved Facebook's acquisition of WhatsApp in 2014. The transaction nevertheless became an important precedent for understanding competition issues surrounding data concentration and platform ecosystems.

The case is particularly relevant to the proposition that a company may possess competitive advantages because of the combination of multiple categories of user data.

Genomic relevance

A genomic company may similarly combine:

• genetic data;
• health records;
• ancestry information;
• medication information;
• clinical-trial information;
• lifestyle information.

The combined dataset can potentially become more valuable than the individual datasets considered separately.

Therefore, merger analysis should examine data complementarity, not merely overlapping databases.

Relevance: Medium.

Case 6: Apple/Shazam, Case M.8788

In Apple/Shazam, the European Commission investigated whether Apple's acquisition of Shazam could create competitive problems because Shazam collected information concerning users of competing music-streaming applications.

The Commission considered whether Apple could gain access to commercially sensitive information about competitors through Shazam's data. It ultimately concluded that the transaction was unlikely to significantly impede effective competition.

Genomic relevance

This case illustrates the information-foreclosure theory.

A genomic platform acquiring another company might gain information about:

• competing laboratories;
• patients using competing services;
• clinical-trial participants;
• researchers;
• pharmaceutical pipelines;
• competitor demand;
• genomic-testing patterns.

The concern may therefore arise not simply from possession of the database, but from the information about competitors revealed by the database.

Relevance: Medium to high.

6. Comparative Case-Law Table

7. Essential-Facility-Type Concerns

A particularly difficult issue is whether a genomic database can constitute an essential facility or indispensable input.

A database may potentially acquire such importance where:

1. competitors cannot reasonably duplicate it;
2. access is indispensable for competing;
3. the database has no realistic substitute;
4. access can technically be provided;
5. refusal substantially impairs downstream competition; and
6. legitimate business or privacy reasons do not justify exclusion.

However, courts generally apply demanding standards before converting competition law into a general right of access.

Therefore, mere usefulness is insufficient.

The stronger the evidence that a particular genomic database is genuinely indispensable, the stronger the potential competition-law argument.

8. Genomic Data and Market Definition

Traditional market definition becomes complicated because genomic databases may support several markets simultaneously.

Potential relevant markets include:

Upstream

• DNA sequencing;
• genomic testing;
• genetic-data collection;
• genomic database services.

Intermediate

• genomic analytics;
• variant interpretation;
• bioinformatics;
• genomic AI.

Downstream

• diagnostics;
• precision medicine;
• pharmaceutical research;
• clinical trials;
• drug discovery.

A merger could therefore generate conglomerate or vertical effects even where the parties have little direct horizontal overlap.

9. The Importance of Dataset Uniqueness

Competition authorities should distinguish between:

Replicable data

Data that competitors can reasonably obtain from:

• public databases;
• customers;
• open research;
• alternative testing services.

and

Non-replicable data

Data involving:

• rare diseases;
• longitudinal patient histories;
• large population cohorts;
• linked genomic/clinical information;
• rare genetic variants;
• family pedigrees;
• longitudinal treatment outcomes.

The second category creates substantially greater competitive concerns.

The Google/Fitbit analysis demonstrates the importance of asking whether the acquired dataset is unique or readily substitutable.

10. Privacy and Competition Law

Genomic data is unusually sensitive.

This produces an important interaction between:

Competition law + data protection + bioethics + healthcare regulation.

Competition authorities should not automatically assume that unlimited data sharing is pro-competitive.

For example, forcing a genomic company to share raw genetic data could itself create:

• privacy risks;
• re-identification risks;
• misuse of genetic information;
• loss of individual control;
• cybersecurity risks.

Therefore, competition remedies must be carefully designed.

11. Possible Competition Remedies

A. Data-silo remedy

The acquiring company may be prohibited from combining the acquired genomic database with its existing datasets.

Google/Fitbit provides an important precedent for this type of approach.

B. Non-discrimination

The dominant database operator may be required to provide equivalent access conditions to competing firms.

C. Interoperability

Authorities may require:

• standard APIs;
• machine-readable formats;
• portability mechanisms;
• technical interoperability.

D. Licensing remedies

A company may be required to license certain datasets on:

• fair;
• reasonable;
• transparent; and
• non-discriminatory

terms, subject to privacy and consent requirements.

E. Divestiture

Where behavioural remedies cannot adequately prevent foreclosure, structural separation may be considered.

The Illumina/GRAIL proceedings demonstrate the importance of divestiture as a structural remedy in a biotechnology concentration.

12. Special Problem of AI and Genomic Databases

The emergence of generative and predictive AI substantially increases the value of genomic datasets.

A firm possessing:

genomic data + clinical data + computing power + AI models

may obtain an advantage that competitors cannot easily reproduce.

This can produce a data-compute-model feedback loop:

Large genomic database
↓
Better AI training
↓
Better diagnosis/drug discovery
↓
More customers/research partnerships
↓
More genomic data
↓
Even better AI

Competition authorities should therefore examine whether a concentration allows the acquiring firm to obtain a self-reinforcing data advantage.

13. Killer-Acquisition Concerns

Genomic startups frequently have:

• low turnover;
• valuable intellectual property;
• promising research;
• large datasets;
• substantial future potential.

Consequently, traditional turnover-based merger thresholds may fail to identify some strategically important acquisitions.

Illumina/GRAIL illustrates this problem particularly clearly: GRAIL's lack of turnover contributed to the jurisdictional controversy surrounding the transaction in the EU. The Court of Justice's 2024 judgment ultimately held that the Commission could not accept the particular Article 22 referral in those circumstances.

The broader policy question remains important:

Should merger-control systems rely solely on current turnover where the principal competitive asset is data, innovation or future technology?

14. Competition Law Test for Genomic Database Concentrations

A useful analytical framework is:

Step 1 — Identify the database

What genomic information is being acquired?

Step 2 — Measure uniqueness

Can competitors reproduce the dataset?

Step 3 — Identify affected markets

Which upstream, downstream or neighbouring markets depend upon it?

Step 4 — Examine market power

Does the acquiring company already possess:

• sequencing power?
• cloud infrastructure?
• AI capability?
• pharmaceutical distribution?
• diagnostic platforms?

Step 5 — Examine foreclosure

Could the company:

• deny access?
• raise prices?
• degrade quality?
• discriminate?
• impose exclusivity?
• combine datasets?

Step 6 — Examine innovation

Would the transaction eliminate an important source of future innovation?

Step 7 — Consider privacy constraints

Can access remedies operate consistently with genomic privacy and consent obligations?

Step 8 — Select remedies

Possible remedies include:

data silo → interoperability → non-discrimination → licensing → divestiture.

15. Application to India

For India, genomic database concentrations can potentially engage the Competition Act, 2002, particularly through merger control and abuse-of-dominance principles.

The Competition Commission of India could potentially examine:

• combinations involving genetic-testing businesses;
• acquisitions of genomic-data companies;
• control over important healthcare datasets;
• denial or discriminatory access to data;
• tying genomic data to sequencing or analytical services;
• exclusive arrangements;
• leveraging of genomic-data dominance into adjacent markets.

The relevant analytical concepts would include:

• relevant market;
• appreciable adverse effect on competition;
• dominance;
• denial of market access;
• discriminatory conditions;
• leveraging;
• vertical restraints; and
• combination effects.

The Indian framework would need to be applied alongside applicable data-protection, healthcare, biomedical research and genetic-data requirements.

16. Key Legal Principles Emerging from the Cases

The six cases collectively demonstrate several principles:

1. Data can constitute a competitive asset

Google/Fitbit demonstrates that competition authorities may examine the competitive implications of combining datasets.

2. Health data can create strategic market power

IQVIA/Propel Media shows how control over critical healthcare datasets can become a central merger concern.

3. Genomic markets require innovation analysis

Illumina/GRAIL demonstrates the importance of innovation and future competition in biotechnology.

4. Data concentration can have vertical effects

A database owner may be able to disadvantage downstream competitors.

5. Data uniqueness matters

The competitive importance of a database depends not merely on its size but on whether rivals can obtain equivalent information.

6. Remedies must address data combination

Data-silo and interoperability remedies may sometimes be more appropriate than traditional price-based remedies.

17. Conclusion

Genomic database concentration represents a developing frontier of competition law. The central competition problem is not simply that one undertaking possesses "a lot of data." The crucial question is whether the concentration gives that undertaking a unique, durable and difficult-to-replicate competitive advantage that can be leveraged across genomic testing, diagnostics, pharmaceuticals, AI and healthcare markets.

The most significant direct precedent is Illumina/GRAIL, while Google/Fitbit and IQVIA/Propel Media demonstrate how competition authorities can assess concentration of health-related datasets. Microsoft/LinkedIn, Facebook/WhatsApp and Apple/Shazam provide broader data-driven merger principles.

Accordingly, a modern genomic merger assessment should move beyond conventional market shares and examine:

data uniqueness + replicability + interoperability + network effects + vertical foreclosure + innovation + AI advantages + privacy constraints.