Continuous Monitoring Enforcement Models

1. Introduction

Continuous Monitoring Enforcement Models refer to regulatory and competition-law systems in which authorities do not rely solely on periodic investigations, complaints, dawn raids, or ex-post enforcement. Instead, they establish mechanisms for the ongoing observation of market behaviour, algorithmic systems, pricing, transactions, access conditions, data practices, compliance indicators, and structural changes.

The concept is particularly important in digital markets because potentially anti-competitive conduct can occur at machine speed and can be modified continuously through software updates, algorithmic pricing, recommendation systems, API changes, ranking adjustments, and automated contracting.

A continuous monitoring model therefore attempts to move enforcement from:

Detection → Investigation → Decision → Remedy

towards a more iterative model:

Monitoring → Detection → Risk assessment → Intervention → Verification → Continued monitoring

It does not, however, mean that a competition authority can automatically presume illegality merely because its monitoring system identifies a suspicious pattern. Legal standards of proof, procedural fairness, causation, proportionality, confidentiality, and judicial review remain important.

2. Meaning of Continuous Monitoring Enforcement

A continuous monitoring enforcement model generally contains five elements:

A. Permanent or periodic data collection

Authorities may monitor:

• prices;
• discounts;
• output;
• market shares;
• switching rates;
• platform access;
• rankings;
• algorithmic changes;
• interoperability;
• licensing conditions;
• exclusivity arrangements;
• acquisition activity;
• complaints;
• consumer behaviour;
• supplier conditions; and
• compliance with previous remedies.

B. Automated or structured detection

Data may be analysed to identify:

• unusual price movements;
• coordinated behaviour;
• exclusionary patterns;
• discriminatory access;
• self-preferencing;
• sudden changes following an acquisition;
• foreclosure indicators;
• margin compression;
• discriminatory algorithmic treatment; or
• repeated breaches of an imposed remedy.

C. Risk-based escalation

Not every detected anomaly becomes an enforcement case.

A monitoring system can classify conduct into:

1. ordinary market variation;
2. conduct requiring clarification;
3. potential compliance concern;
4. potential infringement requiring investigation; and
5. urgent intervention.

D. Intervention

Possible responses include:

• information requests;
• compliance directions;
• behavioural commitments;
• interim measures;
• formal investigation;
• structural remedies;
• fines;
• monitoring trustees;
• independent audits; or
• modification of an existing remedy.

E. Verification

After intervention, the authority continues monitoring whether the undertaking actually complies.

This final stage distinguishes continuous monitoring enforcement from traditional one-time enforcement.

3. Why Continuous Monitoring Has Become Important

3.1 Digital markets change rapidly

Traditional investigations can take years.

Digital platforms, by contrast, can alter:

• algorithms;
• interfaces;
• APIs;
• ranking systems;
• commissions;
• advertising systems;
• search results; and
• access conditions

within hours or days.

Consequently, a remedy that is effective at the date of the decision may become ineffective later.

3.2 Algorithmic conduct can be difficult to observe

An undertaking may not issue an explicit instruction saying:

"Exclude competitors."

Instead, exclusion can emerge through:

• ranking parameters;
• recommendation systems;
• automated pricing;
• default settings;
• data-access restrictions;
• machine-learning optimisation; or
• automated contract enforcement.

Continuous monitoring therefore becomes a method of observing patterns rather than merely individual decisions.

3.3 Structural remedies may require supervision

Where an authority imposes behavioural commitments, the central question becomes:

How does the authority know that the commitment continues to work?

Continuous monitoring provides the answer through:

• reporting obligations;
• independent monitoring;
• data access;
• audits;
• compliance dashboards; and
• periodic reassessment.

4. Legal Foundations

Continuous monitoring can operate under several different legal mechanisms.

A. Competition investigations

Authorities can collect information during investigations and examine market conduct over time.

B. Market investigations and sector inquiries

A regulator may study an entire sector rather than a single undertaking.

C. Merger remedies

Where a merger creates competition concerns, continuing monitoring can ensure compliance with:

• divestiture obligations;
• access commitments;
• firewalls;
• interoperability obligations;
• non-discrimination requirements; or
• information-sharing restrictions.

D. Behavioural commitments

An authority can accept commitments that require continuing compliance.

E. Interim measures

Where immediate harm is possible, continuous monitoring can accompany temporary restrictions while the substantive investigation continues.

F. Digital-market regulation

Modern digital regulation increasingly incorporates:

• reporting;
• auditing;
• transparency;
• data access;
• interoperability;
• risk assessment; and
• supervisory powers.

5. Continuous Monitoring and Article 102 TFEU

Article 102 TFEU prohibits abuse of a dominant position.

Continuous monitoring may become relevant where dominance creates a continuing risk of:

• exclusion;
• discriminatory access;
• tying;
• predatory pricing;
• margin squeeze;
• refusal to supply;
• self-preferencing;
• exploitative conduct; or
• discriminatory conditions.

The important distinction is between monitoring dominance and presuming abuse.

A dominant undertaking is not automatically prohibited from changing its prices, algorithms, contracts, or products.

The authority must still establish the relevant legal infringement.

6. Continuous Monitoring and UK Competition Law

In the UK, continuous monitoring can interact with the enforcement powers of the Competition and Markets Authority (CMA) and sector regulators.

It may be relevant to:

• Chapter II Competition Act 1998 investigations;
• market studies;
• market investigations;
• merger remedies;
• consumer-protection enforcement;
• digital-market supervision; and
• compliance with undertakings.

The model is particularly significant where remedies depend upon continuing behavioural obligations.

7. Six Major Case Laws

Case 1 — United Brands v Commission

Facts

United Brands was found to have abused a dominant position in the banana market, including through discriminatory pricing and other exclusionary practices.

Principle

The case established important principles concerning:

• dominance;
• relevant markets;
• discriminatory conditions; and
• abuse of dominant position.

Relevance to continuous monitoring

The significance for continuous monitoring is that abusive conduct may involve repeated commercial behaviour rather than a single isolated event.

A regulatory system monitoring:

• prices;
• trading conditions;
• customer categories; and
• discriminatory treatment

could identify recurring patterns requiring investigation.

Enforcement lesson

Continuous monitoring should examine patterns of conduct over time, rather than treating every commercial decision as an independent event.

Case 2 — Hoffmann-La Roche v Commission

Facts

The European Commission found that Hoffmann-La Roche had abused its dominant position through loyalty-inducing rebate arrangements.

Principle

The case is foundational for the treatment of exclusionary conduct by dominant firms.

Relevance

A monitoring system could examine:

• rebate structures;
• customer coverage;
• exclusivity effects;
• changes in effective prices;
• duration of arrangements; and
• competitor access.

This illustrates how continuous monitoring can reveal whether apparently ordinary contractual arrangements produce systematic foreclosure.

Enforcement lesson

The relevant object of monitoring is often not merely the contractual clause but its economic operation over time.

Case 3 — Intel v Commission

Facts

Intel concerned rebates and payments associated with the supply of computer processors.

The litigation significantly developed the treatment of conditional rebates by dominant firms and the circumstances in which their exclusionary effects should be assessed.

Principle

The case demonstrates the importance of examining the actual or potential exclusionary effects of conduct rather than relying solely on formal classification.

Relevance to continuous monitoring

A monitoring framework could continuously analyse:

• effective rebate rates;
• customer coverage;
• duration;
• switching possibilities;
• rival access;
• contestable demand; and
• price-cost relationships.

Enforcement lesson

Continuous monitoring should generate evidence capable of supporting a legally meaningful effects analysis rather than merely producing statistical anomalies.

Case 4 — Google Shopping

Facts

The European Commission found that Google had favoured its own comparison-shopping service in search results while demoting competing comparison-shopping services.

The case concerned the interaction between search ranking and competition.

Principle

The case demonstrated how a dominant digital platform's ranking and presentation mechanisms can have competitive consequences.

Relevance to continuous monitoring

This is particularly important for algorithmic enforcement.

A monitoring system could observe:

• ranking positions;
• visibility of competitors;
• traffic allocation;
• algorithmic changes;
• treatment of competing services;
• changes following product updates; and
• differential treatment between internal and external services.

Enforcement lesson

In digital markets, enforcement may need to monitor dynamic system behaviour, rather than only written policies.

Case 5 — Slovak Telekom v Commission

Facts

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

Principle

The litigation dealt with refusal/access-related conduct and the application of competition law to infrastructure-dependent markets.

Relevance

Continuous monitoring becomes particularly useful where competitors depend upon access to infrastructure.

An authority can monitor:

• access requests;
• response times;
• technical conditions;
• pricing;
• capacity;
• interoperability;
• quality of service; and
• discriminatory treatment.

Enforcement lesson

Where market access depends on an infrastructure operator, continuous monitoring can detect gradual foreclosure before it becomes visible through a single dramatic event.

Case 6 — Qualcomm

Facts

The European Commission's Qualcomm proceedings concerned exclusionary payments associated with the supply of baseband chipsets.

Principle

The case illustrates the competition-law significance of conditional commercial incentives in technologically important markets.

Relevance to continuous monitoring

Technology markets can change rapidly.

Monitoring can therefore examine:

• supply contracts;
• conditional payments;
• exclusivity arrangements;
• customer coverage;
• switching;
• rival entry;
• technological developments; and
• changes in contractual incentives.

Enforcement lesson

Continuous monitoring can be valuable in markets where commercial incentives and technological development interact continuously.

8. Additional Important Authorities

Other competition cases provide useful foundations for continuous monitoring.

Michelin I

The Michelin litigation illustrates concerns surrounding loyalty-inducing rebate systems and the importance of examining repeated commercial relationships.

British Airways v Commission

The case concerned incentive schemes and exclusionary effects. It demonstrates the relevance of monitoring incentive structures across a customer base.

Bronner

Bronner is important for understanding exceptional circumstances surrounding refusal of access to infrastructure.

Microsoft

The Microsoft litigation demonstrates the importance of interoperability and access obligations in technology markets.

Deutsche Telekom

The case illustrates how pricing relationships between wholesale and retail levels can generate exclusionary effects and therefore require economically informed analysis.

9. Architecture of a Continuous Monitoring Enforcement Model

A sophisticated model can be represented as follows:

Market Data

↓

Automated Collection

↓

Data Validation

↓

Risk Indicators

↓

Pattern Detection

↓

Human Regulatory Review

↓

Preliminary Assessment

↓

Information Request / Investigation

↓

Intervention

↓

Remedy

↓

Compliance Monitoring

↓

Remedy Verification

↓

Reassessment

↓

Further Intervention if Necessary

This creates an enforcement feedback loop.

10. Continuous Monitoring of Algorithms

Algorithms present a special enforcement challenge.

An authority may monitor:

Pricing algorithms

• price synchronisation;
• rapid parallel price changes;
• personalised prices;
• algorithmic repricing;
• common optimisation objectives.

Ranking algorithms

• ranking changes;
• visibility of rivals;
• preferential treatment;
• demotion patterns.

Recommendation algorithms

• product exposure;
• internal versus external products;
• switching information;
• consumer steering.

Access algorithms

• API permissions;
• authentication requirements;
• rate limits;
• technical degradation;
• interoperability.

The critical legal point is that algorithmic correlation is not automatically proof of an infringement.

11. Continuous Monitoring and Tacit Collusion

One important application concerns algorithmic pricing.

Suppose several firms use pricing systems that repeatedly react to each other's prices.

A monitoring authority could identify:

• unusually rapid price responses;
• persistent price alignment;
• reduced price dispersion;
• repeated reaction patterns; and
• market-wide price movements.

But the authority must distinguish:

lawful independent adaptation

from

concerted or otherwise unlawful coordination.

Continuous monitoring therefore functions as an early-warning system, not necessarily as an automatic liability mechanism.

12. Continuous Monitoring of Merger Remedies

Continuous monitoring is particularly valuable after mergers.

Suppose an authority permits a transaction subject to:

• divestiture;
• access;
• licensing;
• interoperability;
• non-discrimination; or
• information-firewall commitments.

The authority can monitor whether:

1. the remedy has actually been implemented;
2. competitors receive the promised access;
3. prices remain within the agreed parameters;
4. technical functionality is preserved;
5. personnel separation remains effective;
6. information flows comply with the commitment; and
7. market conditions have changed.

This converts merger remedies from static promises into supervised obligations.

13. Monitoring Trustees and Independent Auditors

A continuous monitoring regime may use an independent third party.

The monitor may:

• inspect documents;
• review algorithms;
• examine compliance reports;
• interview personnel;
• test technical systems;
• verify access;
• investigate complaints; and
• report breaches to the authority.

This is especially useful where the authority does not possess sufficient technical resources to inspect a complex digital system itself.

14. Advantages

A. Early detection

Potential violations may be identified before substantial competitive harm occurs.

B. Faster enforcement

Monitoring can reduce the time required to identify relevant evidence.

C. Better digital-market supervision

It is suitable for markets where conditions change continuously.

D. Remedy verification

Authorities can determine whether remedies actually work.

E. Evidence preservation

Continuous data collection may prevent important evidence from disappearing.

F. Reduced reliance on complaints

Authorities need not depend exclusively on competitors or consumers bringing complaints.

15. Risks and Limitations

15.1 False positives

A statistical anomaly does not necessarily constitute unlawful conduct.

For example:

parallel prices ≠ automatically cartel.

15.2 False negatives

Sophisticated firms may design conduct specifically to remain below predefined monitoring thresholds.

15.3 Regulatory overreach

Continuous supervision can create excessive interference with legitimate commercial decisions.

15.4 Confidentiality

Continuous access to commercial data raises issues concerning:

• trade secrets;
• commercially sensitive information;
• personal data; and
• cybersecurity.

15.5 Algorithmic opacity

Authorities themselves may not fully understand a complex machine-learning system.

15.6 Automation bias

Regulators may become overly dependent upon the output of their monitoring systems.

Human legal judgment therefore remains essential.

16. Procedural Safeguards

A legally sustainable model should incorporate:

1. Clear statutory authority

The authority should identify the legal basis for collecting and processing information.

2. Defined monitoring objectives

Monitoring should be linked to legitimate regulatory purposes.

3. Proportionality

The volume and intrusiveness of data collection should correspond to the regulatory objective.

4. Human review

Automated alerts should not automatically determine liability.

5. Right to respond

Undertakings should have an opportunity to explain detected conduct.

6. Confidentiality protections

Trade secrets and protected information require appropriate safeguards.

7. Auditability

Monitoring systems should themselves be capable of being examined.

8. Judicial review

Enforcement decisions must remain subject to applicable legal review.

17. Continuous Monitoring Versus Traditional Enforcement

The two approaches are complementary rather than mutually exclusive.

18. Continuous Monitoring as a Feedback System

The most advanced model can be understood as a regulatory feedback loop:

Observe → Detect → Assess → Intervene → Measure → Adjust

For example:

Observation: platform ranking changes.

↓

Detection: competitor visibility falls significantly.

↓

Assessment: investigate whether the change results from legitimate product improvements or discriminatory self-preferencing.

↓

Intervention: information request or formal investigation.

↓

Remedy: behavioural commitment.

↓

Monitoring: track rankings after the remedy.

↓

Adjustment: modify enforcement if the remedy does not restore effective competition.

This makes competition enforcement an adaptive process.

19. Relationship with Due Process

Continuous monitoring should not eliminate the distinction between:

• suspicion;
• investigation;
• finding of infringement; and
• sanction.

A monitoring system can generate an enforcement signal, but legal liability generally requires the applicable substantive and procedural requirements to be satisfied.

This distinction is particularly important where machine-generated indicators are used as evidence.

20. Emerging Importance in AI Markets

Continuous monitoring is likely to become increasingly relevant to:

• foundation-model markets;
• AI inference services;
• cloud-compute markets;
• AI application stores;
• model marketplaces;
• algorithmic advertising;
• automated procurement;
• AI pricing systems;
• API ecosystems; and
• vertically integrated technology platforms.

For example, a dominant platform could continuously modify an API so that competing applications technically remain available but become progressively slower or more expensive.

A one-time investigation may miss the gradual nature of the restriction.

A continuous monitoring system could compare:

API access → latency → pricing → functionality → competitor usage → algorithmic changes

over time.

21. Key Legal Questions

Continuous monitoring enforcement models raise several important legal questions:

1. When does monitoring become an investigation?
2. How much information can an authority continuously demand?
3. Can algorithmic alerts justify compulsory information requests?
4. Who audits the regulator's monitoring algorithm?
5. How are trade secrets protected?
6. How should false positives be handled?
7. Can behavioural remedies be dynamically modified?
8. What procedural rights must undertakings receive?
9. How should monitoring interact with judicial review?
10. Can continuous monitoring become disproportionate regulatory surveillance?

22. Core Principles Derived from the Case Law

The cases discussed above support several broad principles relevant to continuous monitoring:

Principle 1 — Monitor conduct, not merely formal documents

United Brands and Hoffmann-La Roche illustrate the importance of examining actual commercial behaviour.

Principle 2 — Examine economic effects where legally required

Intel demonstrates the importance of economically meaningful assessment of exclusionary effects.

Principle 3 — Digital systems require technical monitoring

Google Shopping demonstrates why ranking and digital architecture can become competition-law relevant.

Principle 4 — Infrastructure access may require continuing supervision

Slovak Telekom illustrates the importance of access conditions in network-dependent markets.

Principle 5 — Commercial incentives require longitudinal analysis

Qualcomm illustrates how contractual incentives may need to be examined in their market context.

23. Conclusion

Continuous Monitoring Enforcement Models represent a shift from episodic competition enforcement towards adaptive, data-supported supervision. Their principal value lies in markets where competitive conditions can change rapidly and where algorithms, platforms, APIs, pricing systems and contractual arrangements can be modified continuously.

The appropriate model is not:

"Monitor everything and automatically punish anomalies."

It is:

"Continuously observe relevant market signals, identify credible risks, subject those signals to human and legal assessment, intervene where the applicable legal standard is satisfied, and continuously verify the effectiveness of the intervention."

The jurisprudence of United Brands, Hoffmann-La Roche, Intel, Google Shopping, Slovak Telekom and Qualcomm, together with related competition cases, provides the conceptual foundation for this approach: competition enforcement must remain grounded in market context, legally relevant effects, procedural safeguards and proportionality, even when technology permits regulators to observe markets continuously.

Available next action: Create a downloadable PDF file here in this chat containing the findings and recommendations above