Algorithmic Market Governance in Europe

1. Meaning and Scope

Algorithmic market governance refers to the use of algorithms, AI systems, automated pricing tools, data-driven platforms, and computational decision-making to organise, regulate, monitor, influence, or control markets.

It covers both:

1. private algorithmic governance — businesses and digital platforms using algorithms to determine market behaviour; and
2. public algorithmic governance — regulators and public authorities using algorithms to supervise markets and enforce economic regulation.

Examples include:

• algorithmic pricing;
• dynamic pricing;
• automated bidding;
• algorithmic advertising;
• recommendation systems;
• online marketplaces;
• algorithmic credit allocation;
• automated fraud detection;
• supply-chain optimisation;
• algorithmic trading;
• competition-law monitoring;
• automated regulatory supervision;
• market surveillance;
• digital-platform ranking;
• personalised offers;
• automated contract formation.

There is no single European legal cause of action called “algorithmic market governance.” Instead, it is governed through a combination of:

• EU competition law;
• digital-platform regulation;
• consumer law;
• data-protection law;
• financial-market regulation;
• contract law;
• tort/delict law;
• administrative law;
• fundamental rights;
• the EU AI Act;
• the Digital Markets Act;
• the Digital Services Act.

2. Why Algorithms Change Market Governance

Traditional markets generally involved human decisions concerning:

• price;
• supply;
• demand;
• advertising;
• distribution;
• negotiation.

Algorithms can now perform these functions automatically and continuously.

For example:

Ten competing retailers use pricing algorithms that observe competitors' prices and automatically adjust their own prices every few seconds.

This creates difficult legal questions:

• Is there an unlawful agreement?
• Can algorithms collude without explicit communication?
• Is the pricing algorithm facilitating coordination?
• Does the platform manipulate competition?
• Is personalised pricing misleading?
• Is algorithmic ranking discriminatory?
• Is data collection excessive?
• Is the platform abusing market power?

3. Main Legal Framework

A. Article 101 TFEU — Restrictive Agreements

Article 101 prohibits agreements, decisions and concerted practices that restrict competition.

Algorithmic systems can potentially facilitate:

• price fixing;
• market sharing;
• output restrictions;
• bid rigging;
• information exchange;
• coordination between competitors.

The fact that the coordination occurs through software does not necessarily prevent Article 101 from applying.

4. Article 102 TFEU — Abuse of Dominance

Algorithms become particularly important where a dominant company controls:

• search;
• advertising;
• online marketplaces;
• app stores;
• cloud infrastructure;
• data;
• digital distribution.

Potential abuses include:

• self-preferencing;
• discriminatory access;
• tying;
• exclusionary ranking;
• predatory pricing;
• discriminatory pricing;
• refusal of access;
• exploitation of data advantages.

5. Digital Markets Act

The Digital Markets Act (DMA) adds specific obligations for designated gatekeepers.

It addresses certain practices involving:

• ranking;
• self-preferencing;
• data combination;
• interoperability;
• platform access;
• advertising data;
• business-user relationships.

The DMA is important because traditional competition law often requires lengthy case-by-case analysis, while the DMA imposes specified obligations on designated gatekeepers.

6. Algorithmic Pricing

Algorithmic pricing is one of the most important areas of algorithmic market governance.

Algorithms can:

• monitor competitors' prices;
• predict demand;
• change prices automatically;
• personalise prices;
• identify consumers likely to pay more;
• respond to competitor behaviour.

Algorithmic pricing is not automatically unlawful.

The legal problem arises when algorithms facilitate:

• collusion;
• exclusion;
• discrimination;
• misleading pricing;
• exploitation of consumers.

7. Algorithmic Collusion

A major competition-law question is:

Can competitors violate competition law through algorithms without directly communicating with each other?

There are several possibilities.

Explicit algorithmic agreement

Competitors agree to use a common pricing algorithm to maintain prices.

This is relatively straightforward under Article 101.

Hub-and-spoke coordination

A platform or intermediary may coordinate pricing among competitors.

Tacit algorithmic coordination

Algorithms independently observe market conditions and repeatedly respond to each other.

This is legally more difficult because EU competition law generally requires the relevant legal conditions for an agreement or concerted practice rather than merely parallel behaviour.

Therefore:

Parallel algorithmic pricing does not automatically establish an Article 101 infringement.

8. Eturas — C-74/14

Case: Eturas UAB and Others v Lietuvos Respublikos konkurencijos taryba, C-74/14
CJEU, 21 January 2016

This is one of the most important European cases for algorithmic market governance.

The case concerned an online booking system used by travel agencies.

The system administrator sent a message concerning a restriction on discounts available through the online platform.

Importance

The CJEU examined whether the participating undertakings could be considered involved in a concerted practice where the platform's technical system facilitated the restriction.

Algorithmic significance

The case is highly relevant because it demonstrates that:

Digital infrastructure can be a mechanism through which competitors coordinate market conduct.

An undertaking may face competition-law consequences where it knew, or could reasonably be expected to know, about the anti-competitive communication and continued participating in the system.

Relevance: Very high.

9. United Brands — 27/76

Case: United Brands Company and United Brands Continentaal BV v Commission, Case 27/76
CJEU, 14 February 1978

The case is a foundational EU competition-law authority concerning:

• dominance;
• excessive pricing;
• market power;
• abusive conduct.

Algorithmic relevance

Algorithms can make pricing and market-power analysis substantially more sophisticated.

For example, an algorithmically determined price may potentially be examined for:

• excessive pricing;
• discriminatory pricing;
• exclusionary pricing.

The algorithm does not itself eliminate the underlying Article 102 principles.

Relevance: Foundational/analogical.

10. Google Shopping — C-48/22 P

Case: Google and Alphabet v Commission, C-48/22 P
CJEU, 10 September 2024

This modern competition case concerns Google's conduct in comparison-shopping services.

The Court upheld the essential finding that Google had abused its dominant position through conduct favouring its own comparison-shopping service in search results.

Algorithmic significance

Search ranking is itself an algorithmic market-governance mechanism.

An algorithm can determine:

• which competitor is visible;
• which product appears first;
• which marketplace receives traffic;
• which business reaches consumers.

The case demonstrates that algorithmic ranking can become the subject of Article 102 scrutiny.

Relevance: Extremely high.

11. Google Shopping and Self-Preferencing

The case is particularly important because a dominant platform can act as both:

• market infrastructure; and
• competitor.

That creates the possibility that the platform's ranking algorithm gives preferential treatment to its own services.

For example:

A dominant marketplace ranks its own products above competing sellers regardless of objectively relevant ranking factors.

The legal issue may involve:

• exclusionary effects;
• discrimination;
• self-preferencing;
• foreclosure;
• reduced consumer choice.

12. Intel — C-413/14 P

Case: Intel Corporation v European Commission, C-413/14 P
CJEU, 6 September 2017

The case concerned rebates by a dominant undertaking.

The CJEU clarified the importance of analysing the actual competitive effects of certain rebate practices where the Commission had relied upon an as-efficient-competitor analysis.

Algorithmic significance

Modern platforms can use algorithms to determine:

• discounts;
• rebates;
• advertising incentives;
• preferential placement;
• commission structures.

Algorithmic implementation does not remove the need for appropriate competition analysis.

Relevance: High by analogy.

13. Amazon and Marketplace Governance

Digital marketplaces raise complex algorithmic governance questions because the platform may simultaneously:

• host third-party sellers;
• collect their data;
• rank their products;
• advertise competing products;
• sell its own products.

Algorithms can therefore influence:

• visibility;
• ranking;
• pricing;
• consumer recommendations;
• access to customers.

These issues are increasingly addressed through both competition law and the DMA.

14. Amazon EU Sàrl — C-815/18 P

EU litigation involving Amazon and competition/market regulation illustrates the broader tension between platform structure, market access and regulatory oversight.

More broadly, Amazon-related EU competition and consumer jurisprudence demonstrates that digital platforms cannot assume that their technological architecture places them outside ordinary market regulation.

For algorithmic market governance, the key issue remains the economic function of the algorithm, not its technical form.

15. Meta Platforms — C-252/21

Case: Meta Platforms Inc. and Others v Bundeskartellamt, C-252/21
CJEU, 4 July 2023

This is an important case at the intersection of:

• competition law;
• personal data;
• behavioural advertising;
• platform power.

The CJEU considered the relationship between GDPR compliance and competition-law assessment.

Algorithmic significance

Platforms often use algorithms to combine:

• platform activity;
• browsing behaviour;
• information from other services;
• advertising data.

This enables highly sophisticated behavioural profiling.

Principle

Competition authorities may take GDPR compliance into account when assessing abuse of dominance, while competition law and data-protection law remain distinct legal regimes.

Relevance: Very high.

16. Google Spain — C-131/12

Case: Google Spain SL and Google Inc. v AEPD and Mario Costeja González, C-131/12
CJEU, 13 May 2014

The case concerned search-engine processing of personal information.

Market-governance significance

Search engines exercise substantial influence over:

• information visibility;
• advertising;
• consumer attention;
• business traffic.

The case shows that algorithmic market power is not only about prices.

Control over information and visibility can itself have enormous economic significance.

Relevance: High.

17. Google v CNIL — C-507/17

Case: Google LLC v CNIL, C-507/17
CJEU, 24 September 2019

The case concerned the territorial scope of de-referencing obligations.

Algorithmic market significance

Search algorithms operate globally, while legal obligations are frequently territorial.

The case illustrates the difficult question:

How far can one jurisdiction regulate a global algorithmic system?

This is particularly relevant for:

• global platforms;
• search engines;
• international advertising;
• cross-border marketplaces.

Relevance: High.

18. Ryanair v PR Aviation — C-30/14

Case: Ryanair Ltd v PR Aviation BV, C-30/14
CJEU, 15 January 2015

The case concerned database use and contractual restrictions.

Algorithmic significance

Modern markets depend heavily upon:

• databases;
• APIs;
• price information;
• product information;
• aggregated datasets.

Contractual control over market data can therefore affect competition and algorithmic market access.

Relevance: Analogical but useful.

19. Huawei Technologies v ZTE — C-170/13

Case: Huawei Technologies Co. Ltd v ZTE Corp. and ZTE Deutschland GmbH, C-170/13
CJEU, 16 July 2015

The case concerned standard-essential patents and competition law.

Algorithmic relevance

Modern AI and digital markets depend on technical standards and interoperability.

Where a company controls technology essential to market participation, legal questions can arise concerning:

• access;
• licensing;
• interoperability;
• market foreclosure.

The principle is particularly relevant to AI ecosystems where technical standards become essential infrastructure.

Relevance: Analogical.

20. UPC Telekabel — C-314/12

Case: UPC Telekabel Wien GmbH v Constantin Film Verleih GmbH and Wega Filmproduktionsgesellschaft mbH, C-314/12
CJEU, 27 March 2014

The CJEU considered injunctions requiring internet access providers to block access to copyright-infringing websites.

Algorithmic significance

Digital-market governance requires balancing:

• intellectual property;
• business freedom;
• information access;
• technological implementation.

It demonstrates that technical market controls can have fundamental-rights consequences.

Relevance: Moderate to high.

21. SCHUFA — C-634/21

Case: SCHUFA Holding AG (Scoring), C-634/21
CJEU, 7 December 2023

Although primarily a data-protection case, SCHUFA is also highly relevant to algorithmic market governance.

Credit scores can determine access to:

• loans;
• financial services;
• housing;
• commercial opportunities.

Importance

Where an algorithmic score plays a decisive role in a subsequent decision, Article 22 GDPR can become relevant.

Market-governance significance

A private scoring system can become a form of private economic infrastructure.

Its errors can influence people's participation in markets.

Relevance: Very high.

22. Algorithmic Consumer Governance

Algorithms increasingly govern consumer markets through:

• personalised advertising;
• personalised prices;
• product ranking;
• recommendations;
• targeted discounts;
• automated credit offers.

Consumer-protection law becomes relevant where algorithms:

• mislead consumers;
• hide material information;
• exploit vulnerabilities;
• manipulate purchasing decisions;
• create unfair commercial practices.

23. Personalised Pricing

Personalised pricing occurs when an algorithm estimates the price different consumers are likely to accept.

For example:

Consumer A sees €50 while Consumer B sees €75 for the same product.

Personalised pricing is not automatically unlawful.

But potential issues include:

• transparency;
• discrimination;
• consumer manipulation;
• unfair commercial practices;
• data-protection violations;
• competition concerns.

24. Algorithmic Ranking

Ranking algorithms can determine:

• which seller appears first;
• which hotel is recommended;
• which app is visible;
• which advertisement receives attention.

This can create a powerful form of non-price market control.

The Google Shopping litigation is particularly significant because it demonstrates how search ranking can affect competitive conditions.

25. Self-Preferencing

Self-preferencing occurs where a platform gives its own products or services preferential treatment over competitors.

Example:

A marketplace operates a ranking algorithm that consistently places its own products above competing sellers.

Potential legal frameworks include:

• Article 102 TFEU;
• Digital Markets Act;
• national competition law;
• consumer law.

26. Algorithmic Advertising

Advertising algorithms can decide:

• which user sees an advertisement;
• how much an advertiser pays;
• where the advertisement appears;
• which audience receives it;
• how often it is displayed.

The system may use:

• browsing history;
• location;
• purchases;
• demographics;
• behavioural predictions.

This creates overlap between:

competition law + consumer law + GDPR + platform regulation.

27. Algorithmic Market Surveillance by Government

Algorithmic governance is not limited to companies.

Regulators may use algorithms to detect:

• insider trading;
• market manipulation;
• tax evasion;
• price manipulation;
• cartel behaviour;
• suspicious transactions;
• financial misconduct.

Such systems can process huge volumes of transactions.

Public-sector use creates additional requirements concerning:

• legality;
• proportionality;
• transparency;
• administrative fairness;
• data protection;
• effective judicial review.

28. Financial-Market Algorithms

Financial markets provide some of the clearest examples of algorithmic governance.

Algorithms can perform:

• high-frequency trading;
• automated portfolio management;
• risk assessment;
• fraud detection;
• market surveillance;
• credit scoring.

Potential claims include:

• market manipulation;
• negligent programming;
• inadequate controls;
• discriminatory financial access;
• cybersecurity failures.

29. Algorithmic Market Governance and Fundamental Rights

Market algorithms can affect:

Privacy

Through extensive behavioural profiling.

Equality

Through discriminatory pricing or ranking.

Freedom to conduct a business

Where regulation restricts algorithmic operations.

Freedom of expression

Where platform algorithms determine communication visibility.

Property rights

Where algorithmic systems substantially affect valuable digital assets.

European law therefore requires balancing economic interests with fundamental rights.

30. Freedom to Conduct Business

The EU Charter's Article 16 protects the freedom to conduct a business.

The right is not absolute.

Sky Österreich — C-283/11

Case: Sky Österreich GmbH v Österreichischer Rundfunk, C-283/11
CJEU, 22 January 2013

The CJEU recognised the importance of the freedom to conduct a business while accepting that the right can be subject to proportionate regulation.

Algorithmic significance

Rules regulating:

• algorithmic pricing;
• platform design;
• AI systems;
• data use;
• ranking algorithms

can interfere with commercial freedom, but such restrictions can be legitimate if proportionate.

31. Algorithmic Market Governance and AI Regulation

The EU AI Act adds another layer.

For relevant high-risk systems, organisations may need to address:

• risk management;
• data quality;
• technical documentation;
• record keeping;
• human oversight;
• accuracy;
• robustness;
• cybersecurity;
• post-market monitoring.

Market participants therefore increasingly face technology governance duties in addition to traditional competition obligations.

32. The “Black Box” Problem

A major litigation problem is that the affected party may not know:

• what variables were used;
• how the score was generated;
• why the algorithm ranked one business above another;
• why a consumer received a particular price;
• why an account was flagged;
• why a competitor was excluded.

This creates an information asymmetry between:

algorithm operator ↔ affected market participant.

Data-protection, platform-transparency and procedural rules increasingly attempt to reduce this asymmetry.

33. Algorithmic Market Governance and Competition Compliance

Companies should not assume:

“The algorithm made the decision, so there was no human agreement.”

Competition authorities may examine:

• communications between firms;
• algorithm specifications;
• platform instructions;
• common pricing tools;
• implementation arrangements;
• knowledge of coordinated behaviour;
• economic effects.

Eturas is particularly important because it demonstrates how digital infrastructure can facilitate concerted practices.

34. Liability of Different Actors

Algorithm developer

Potential responsibility for:

• defective design;
• inadequate safeguards;
• foreseeable anti-competitive functionality.

Platform operator

Potential responsibility for:

• ranking;
• self-preferencing;
• discriminatory access;
• data exploitation;
• market foreclosure.

Market participant

Potential responsibility for:

• coordinating with competitors;
• using an algorithm for unlawful pricing;
• exploiting algorithmic market power.

Public regulator

Potential responsibility for unlawful regulatory decisions or disproportionate algorithmic enforcement.

35. Typical Algorithmic Market Claims

1. Algorithmic cartel claim

Competitors allegedly use algorithms to coordinate prices.

2. Self-preferencing claim

Dominant platform allegedly ranks its own products preferentially.

3. Algorithmic exclusion claim

Algorithm allegedly prevents competitors from reaching consumers.

4. Algorithmic discrimination claim

Different businesses or consumers receive systematically different treatment.

5. Data exploitation claim

Dominant platform uses data obtained from business users to compete against them.

6. Algorithmic consumer manipulation claim

Personalisation allegedly exploits consumer vulnerabilities.

7. Automated market-surveillance claim

A regulator allegedly makes an unlawful decision based upon automated market-risk classification.

36. Evidence in Algorithmic Market Litigation

Important evidence may include:

• source code;
• model documentation;
• algorithm specifications;
• pricing logs;
• ranking records;
• API documentation;
• platform contracts;
• communications;
• internal emails;
• audit reports;
• competition-law compliance policies;
• data-processing records;
• statistical evidence;
• market-share data;
• A/B testing results.

Economic experts may also examine:

• price effects;
• foreclosure;
• consumer welfare;
• market definition;
• counterfactual scenarios.

37. Causation and Economic Effects

Algorithmic competition claims often require a sophisticated analysis of causation.

For example:

Algorithmic ranking → lower competitor visibility → fewer customers → reduced sales → market foreclosure

or:

Algorithmic coordination → higher prices → consumer overpayment → economic damage

The claimant may therefore require:

• econometric analysis;
• market data;
• counterfactual analysis;
• technical examination.

38. Remedies

Potential remedies include:

Competition remedies

• fines;
• behavioural commitments;
• access obligations;
• non-discrimination requirements;
• interoperability;
• cessation of abusive conduct.

Private-law remedies

• damages;
• injunctions;
• contract remedies.

Regulatory remedies

• corrective orders;
• compliance programmes;
• algorithmic modifications;
• monitoring.

Consumer remedies

• refunds;
• correction;
• compensation;
• cessation of misleading practices.

39. Consolidated Case-Law Table

40. Direct vs Analogical Case Law

Strongly algorithmic/digital

• Eturas
• Google Shopping
• Meta Platforms
• SCHUFA
• Google Spain
• Google v CNIL

Foundational competition authorities

• United Brands
• Intel

Supporting digital-market authorities

• Ryanair v PR Aviation
• Huawei v ZTE
• UPC Telekabel
• Sky Österreich

This distinction is important because European competition law has a much longer history than AI-specific regulation.

41. Practical Legal Test

An algorithmic market-governance dispute can be analysed through ten questions:

1. What market is affected?

Identify the relevant product/service and geographic market.

2. What algorithm is involved?

Pricing? Ranking? Advertising? Recommendation? Scoring?

3. Who controls it?

Platform, competitor, regulator, intermediary or third-party provider?

4. Is the operator dominant?

If yes, Article 102 and/or DMA concerns may arise.

5. Is there coordination?

Examine Article 101.

6. Does the algorithm discriminate?

Consider consumers and competing businesses.

7. Does it exploit personal data?

Consider GDPR.

8. Does it manipulate or mislead consumers?

Consider consumer law.

9. Does it affect fundamental rights?

Consider the Charter and ECHR.

10. What remedy is appropriate?

Damages, injunction, regulatory order, algorithmic modification, access or structural remedy?

42. Central Legal Principles

Principle 1 — Algorithms are not outside competition law

Digital implementation does not immunise anti-competitive conduct.

Principle 2 — Digital infrastructure can facilitate coordination

Eturas is especially important.

Principle 3 — Ranking itself can be economically significant

Google Shopping demonstrates the importance of algorithmic visibility.

Principle 4 — Data can be a source of market power

Meta Platforms illustrates the interaction between data processing and competition law.

Principle 5 — Scores can become economic infrastructure

SCHUFA demonstrates the importance of algorithmic scoring.

Principle 6 — Market power is not limited to price

Control over:

• data;
• rankings;
• visibility;
• recommendations;
• access

can have competitive consequences.

Principle 7 — Regulation must remain proportionate

Sky Österreich illustrates the balancing between commercial freedom and legitimate regulation.

43. Overall European Position

European algorithmic market governance is developing toward a multi-layered regulatory model.

The traditional competition-law framework remains applicable:

Article 101 TFEU → anti-competitive agreements and concerted practices

Article 102 TFEU → abuse of dominance

But it is increasingly supplemented by:

Digital Markets Act → platform/gatekeeper obligations

Digital Services Act → platform transparency and systemic-risk obligations

GDPR → data and profiling controls

AI Act → risk-based AI governance

Consumer law → protection against manipulation and unfair practices

Fundamental rights → proportionality and effective remedies

The most important cases are Eturas, Google Shopping, Meta Platforms, SCHUFA, Google Spain, and Google v CNIL, supported by foundational authorities such as United Brands, Intel, Huawei v ZTE, UPC Telekabel, Ryanair v PR Aviation, and Sky Österreich.

Core formula

Algorithm + market power/coordination + economic effect + applicable legal duty + anti-competitive or unlawful conduct + causation/harm → potential regulatory or civil liability.

The central European principle is therefore:

An algorithm is a technological mechanism, not a legal shield.

Where an algorithm controls prices, rankings, access, data, advertising, credit or market visibility, European law can examine the economic function and real-world effects of that system rather than accepting its technical complexity as a reason to avoid legal scrutiny.