Competition Law and Future Regulation of Signal-Driven Markets

1. Introduction

Signal-driven markets are markets in which competitive decisions are increasingly shaped by signals generated from data, algorithms, platforms, ranking systems, forecasts, recommendations, behavioural information, real-time prices, capacity indicators, or other machine-readable information.

A traditional market works through relatively observable competitive variables such as price, quantity, quality and output. In a signal-driven market, firms may compete by responding to a much larger information environment:

• real-time prices of competitors;
• algorithmic pricing recommendations;
• search rankings and visibility;
• consumer demand forecasts;
• inventory and capacity signals;
• bidding signals;
• platform recommendations;
• supplier and customer data;
• machine-generated forecasts;
• API signals;
• signals about future pricing intentions; and
• signals generated indirectly by a dominant intermediary.

The central competition-law question is therefore changing from:

“Did competitors expressly agree on a price?”

to a more difficult question:

“Did the market's information architecture reduce independent decision-making sufficiently to produce or facilitate anticompetitive coordination?”

This is particularly important because an algorithm can coordinate market behaviour without competitors communicating directly with one another. Competition authorities therefore increasingly examine information exchanges, common algorithms, platform design, self-preferencing, data concentration and algorithmic coordination.

2. Meaning of Signal-Driven Markets

A signal is information capable of influencing another market participant's decision.

Examples include:

The problem becomes particularly acute where many firms receive the same signal from a common technological intermediary.

3. Why Signal-Driven Markets Create New Competition Problems

A. Increased market transparency

Transparency is not always pro-competitive.

If firms can instantly observe each other's prices and adjust automatically, competition may become less independent.

B. Reduced reaction time

Traditional competition may involve delays between:

1. a competitor changing price;
2. discovering the change;
3. analysing it;
4. deciding whether to respond.

Algorithms can reduce this process to milliseconds.

C. Common information intermediaries

A third-party platform may collect information from competing firms and generate recommendations for all of them.

This raises the possibility that the intermediary becomes a coordination infrastructure.

D. Predictive signals

Future-oriented signals are more problematic than purely historical information because they can reveal strategic intentions.

E. Feedback loops

Signal-driven markets can produce:

Firm data → algorithm → market signal → competitor response → new data → algorithmic adjustment

This can create a self-reinforcing competitive environment.

4. Existing Competition-Law Framework

Signal-driven markets do not necessarily require an entirely new competition-law doctrine.

Existing doctrines can potentially address them through:

4.1 Cartel and concerted-practice rules

Competition law can prohibit:

• price fixing;
• market sharing;
• output restriction;
• bid rigging;
• exchange of competitively sensitive information;
• concerted practices; and
• indirect coordination.

4.2 Abuse of dominance

A dominant signal provider may potentially engage in:

• discriminatory access;
• self-preferencing;
• exclusionary algorithmic design;
• refusal to provide essential data;
• tying;
• leveraging;
• discriminatory ranking; or
• exploitation of informational advantages.

4.3 Merger control

Acquisitions can concentrate:

• datasets;
• prediction capabilities;
• recommendation infrastructure;
• customer information;
• APIs;
• cloud infrastructure; and
• algorithmic capabilities.

4.4 Unilateral conduct

A dominant firm may manipulate signals in ways that disadvantage competitors while appearing to compete merely through technology.

5. Six Important Case Laws

Case 1: T-Mobile Netherlands BV v Raad van bestuur van de Nederlandse Mededingingsautoriteit

Court: Court of Justice of the European Union
Case: C-8/08
Principle: Information exchange and concerted practice

The case concerned meetings between mobile telecommunications operators where commercially sensitive information was exchanged.

The CJEU emphasised that an exchange of information can reduce uncertainty concerning competitors' future conduct and therefore potentially constitute a concerted practice.

Importance for signal-driven markets

The principle is highly relevant to modern algorithmic systems.

A digital market does not necessarily need an explicit agreement saying:

“We will charge the same price.”

If firms deliberately exchange or receive strategically important information that reduces uncertainty about their future competitive conduct, competition law may intervene.

In a signal-driven market, the equivalent signal could be:

• intended price;
• planned capacity;
• future discounts;
• expected output;
• algorithmic pricing strategy.

The key issue becomes whether the signal reduces strategic uncertainty between competitors.

6. Case 2: Eturas UAB and Others v Lietuvos Respublikos konkurencijos taryba

Court: CJEU
Case: C-74/14
Subject: Common electronic platform and algorithmic pricing signals

This is one of the most significant cases for digital signal-driven competition.

Several travel agencies used a common electronic booking system. A message sent through the system informed participants about a limitation concerning discounts.

The CJEU considered when knowledge of such a message, combined with continued participation in the system, could support an inference of participation in a concerted practice.

Competition-law significance

The case demonstrates that a digital platform itself can become the channel through which competitive signals are transmitted.

The important lesson for future markets is:

Coordination does not necessarily require a traditional meeting between competitors.

A platform notification, algorithmic instruction, common dashboard or automated system may perform a similar economic function.

Future application

This principle could become important for:

• hotel pricing systems;
• ride-hailing platforms;
• e-commerce marketplaces;
• airline pricing;
• food-delivery platforms;
• advertising exchanges; and
• property-rental algorithms.

7. Case 3: Google Shopping

General Court: Google LLC and Alphabet Inc. v European Commission, Case T-612/17
CJEU appeal: Case C-48/22 P

The Google Shopping litigation concerned Google's treatment of comparison-shopping services within its search results.

The European Commission found that Google favoured its own comparison-shopping service in search results. The General Court upheld the essential finding, and the CJEU's 2024 judgment addressed the legal analysis of potential foreclosure effects and the relationship between Google's general search service and specialised comparison-shopping services.

Relevance to signal-driven markets

Search ranking is itself a market signal.

A ranking signal tells consumers:

“This result is more relevant or valuable than another result.”

If a dominant platform controls the signal architecture, it can potentially influence the competitive position of downstream businesses.

This creates a distinction between:

Competition over products

and

Competition over the signals that determine which products consumers see.

Future regulatory concern

Competition authorities may therefore increasingly investigate:

• ranking manipulation;
• recommendation bias;
• self-preferencing;
• preferential visibility;
• suppression of rival signals;
• discriminatory recommendation systems.

8. Case 4: United States v RealPage, Inc.

Court: U.S. District Court for the Middle District of North Carolina
Filed: 2024

RealPage is particularly significant for the emerging law of algorithmic coordination.

The U.S. Department of Justice alleged that competing landlords supplied competitively sensitive information to RealPage's revenue-management software and received algorithmically generated rental-price recommendations. The complaint alleged violations of Sections 1 and 2 of the Sherman Act.

The later proceedings continued to focus on the use of competitors' sensitive information, algorithmic pricing and the alignment of pricing decisions. In November 2025, the DOJ proposed a settlement requiring restrictions on the use of competitors' non-public information and removal or redesign of certain pricing-alignment features, subject to court approval.

Why RealPage matters

This is an important example of a potential shift from:

human-to-human cartel coordination

toward:

information-to-algorithm-to-market coordination.

The competitors do not necessarily need to call each other and agree on a price.

Instead:

Competitor data → common algorithm → pricing recommendation → competitor response

may produce a similar competitive concern.

Legal significance

Future competition law will need to determine:

• when algorithmic recommendations constitute coordination;
• whether use of common software creates an agreement or concerted practice;
• what degree of human involvement is necessary;
• whether firms remain genuinely independent;
• whether competitively sensitive information can lawfully be supplied to an intermediary.

9. Case 5: United States v Airline Tariff Publishing Co.

Court: U.S. Department of Justice enforcement proceeding
Subject: Airline pricing information and electronic communication

The Airline Tariff Publishing matter involved sophisticated electronic dissemination of airline pricing information.

The DOJ challenged practices involving communication of fare information that could facilitate coordination among airlines.

Relevance

Although technologically older than today's AI systems, the case is conceptually important because it illustrates that communication infrastructure can itself facilitate coordination.

The technological progression is:

telephone → electronic bulletin → database → API → algorithm → autonomous pricing system.

The underlying competition-law concern can remain similar:

Does the information system facilitate independent competition, or does it make coordinated conduct easier?

10. Case 6: United States v Apple Inc.

Court: U.S. District Court for the Southern District of New York
Subject: Platform power, information architecture and exclusion

The Apple litigation concerns alleged exclusionary conduct involving Apple's control over the iPhone ecosystem.

Although it is not an algorithmic-collusion case in the narrow sense, it is relevant to signal-driven markets because dominant digital ecosystems can control:

• access;
• distribution;
• interoperability;
• visibility;
• technical permissions;
• user information; and
• competitive pathways.

The case demonstrates the broader competition-law problem of control over the architecture through which market signals and consumer choices flow.

Future significance

A platform may possess substantial competitive power even when it does not directly set competitors' prices.

Control over:

• recommendation;
• discoverability;
• interoperability;
• defaults;
• data;
• APIs;
• rankings

may influence competition indirectly.

11. Case 7: United States v Google LLC — Search Distribution

The U.S. Google search litigation is also relevant to signal-driven markets.

The competition concern involved agreements and practices concerning distribution of general search services.

Search engines do not merely sell a product. They determine the information environment through which consumers encounter competing products and services.

Consequently, control over search distribution and defaults can affect:

• consumer attention;
• traffic;
• advertising;
• downstream competition;
• data accumulation; and
• future algorithmic improvement.

This illustrates the concept of signal power: controlling the information pathway can be competitively significant even where the controlled signal has no monetary price.

12. The Central Concept: Signal Power

Future competition law may increasingly recognise a concept that can be called signal power.

Signal power is the ability of an undertaking to influence competitive outcomes through control over information that other market participants rely upon.

It can arise from control over:

1. price signals;
2. ranking signals;
3. recommendation signals;
4. availability signals;
5. quality signals;
6. risk signals;
7. demand forecasts;
8. consumer-intent signals;
9. advertising signals;
10. algorithmic predictions.

A company need not own the entire market to possess substantial signal power.

13. Signal-Driven Markets and Tacit Coordination

One of the most difficult issues is tacit coordination.

Suppose four firms use algorithms that observe market prices.

Firm A raises price.

The algorithms of Firms B, C and D immediately detect this signal and respond.

A then observes the responses.

The market may converge toward higher prices without:

• an explicit cartel;
• a meeting;
• an email;
• a telephone call; or
• a written agreement.

Competition law traditionally distinguishes between independent adaptation and concerted conduct.

The challenge is determining where the boundary lies.

14. Human Intention vs Algorithmic Conduct

Traditional cartel law often examines human behaviour.

Future cases may require examination of:

• algorithm design;
• training data;
• objective functions;
• pricing constraints;
• system instructions;
• model architecture;
• feedback mechanisms;
• software contracts;
• data inputs;
• recommendation outputs;
• internal communications concerning algorithm deployment.

This means competition authorities may increasingly need technical evidence, not merely conventional documentary evidence.

15. Competitively Sensitive Information

The following information may require particular scrutiny:

High-risk information

• future prices;
• future discounts;
• future output;
• capacity plans;
• strategic business plans;
• customer-specific information;
• future bidding intentions.

Potentially lower-risk information

• genuinely historical data;
• aggregated data;
• sufficiently anonymised information;
• publicly available information.

However, aggregation and anonymisation must be assessed economically rather than assumed to eliminate competition concerns.

16. Common Algorithm Problem

A major future issue is the use of a common pricing algorithm by competitors.

For example:

Competitor A + Competitor B + Competitor C
↓
Common pricing software
↓
Common data pool
↓
Algorithmic recommendations

The legal question is not simply whether the algorithm belongs to an independent software company.

Authorities may ask:

1. What data does it receive?
2. Is the data competitively sensitive?
3. Does the algorithm observe competitors' current conduct?
4. Does it recommend future prices?
5. Does it discourage price reductions?
6. Does it align competitors' conduct?
7. Can firms independently override the recommendations?
8. Did firms understand the coordination effects?

The RealPage litigation illustrates why these questions are becoming important.

17. Signal Manipulation by Dominant Platforms

A dominant platform can potentially manipulate signals without directly imposing a price.

For example:

Search engine

→ ranking signal

→ consumer attention

→ traffic

→ sales

Similarly:

App store

→ recommendation signal

→ downloads

→ developer success

Or:

Marketplace

→ seller-ranking signal

→ visibility

→ transactions.

Thus, competition authorities may need to assess non-price competitive parameters.

18. Data as a Competitive Signal

Data has two distinct competition functions.

First: Data as an input

Data may be required to operate an algorithm.

Second: Data as a signal

Data may reveal:

• consumer preferences;
• demand;
• competitor behaviour;
• inventory;
• pricing;
• purchasing patterns.

A dominant firm possessing a superior data stream may therefore have a competitive advantage that compounds over time.

19. Feedback Loops

A particularly important feature of signal-driven markets is the feedback loop.

Basic model

More users

↓

More data

↓

Better prediction

↓

Better recommendations

↓

More users

↓

More data

This can produce a reinforcing advantage.

A second loop can occur between competitors:

Competitor data

↓

Common algorithm

↓

Recommendation

↓

Competitor action

↓

New market data

↓

Improved recommendation

The second loop raises the possibility of algorithmic coordination.

20. Future Regulation

Future regulation is likely to develop around several principles.

A. Algorithmic transparency

Authorities may require dominant or high-risk platforms to provide sufficient information about:

• decision parameters;
• ranking systems;
• pricing logic;
• recommendation structures;
• data inputs.

This does not necessarily mean public disclosure of source code.

B. Auditability

Businesses may be required to maintain:

• algorithm logs;
• version histories;
• training-data records;
• model changes;
• decision records;
• human overrides.

This would make competition investigations more effective.

C. Restrictions on sensitive data

Future regulation may limit the transmission of:

• future pricing;
• future capacity;
• future output;
• strategic bidding information

through common algorithmic intermediaries.

D. Algorithmic firewalls

Competitors could be required to maintain technological separation preventing an algorithm from receiving or transmitting certain competitively sensitive information.

E. Independent decision-making requirements

Competition regulation may increasingly focus on whether each competitor retains genuine independent control over:

• pricing;
• output;
• discounts;
• supply;
• bids.

21. Ex Ante Regulation

Traditional competition law is often ex post:

Conduct occurs → investigation → infringement decision → remedy.

Signal-driven markets may require more ex ante intervention.

Possible measures include:

• mandatory algorithmic-risk assessments;
• interoperability requirements;
• data-access rules;
• audit obligations;
• restrictions on self-preferencing;
• limits on discriminatory ranking;
• information-sharing safeguards.

This is particularly relevant where network effects and feedback loops can rapidly reinforce market power.

22. Competition Law and AI Agents

The emergence of autonomous AI agents creates an additional problem.

Imagine:

AI Agent A negotiates for Firm A
AI Agent B negotiates for Firm B
AI Agent C negotiates for Firm C

If the agents learn from market signals and adapt their conduct automatically, the resulting market may become highly coordinated without conventional human communication.

Future legal questions include:

• Who is responsible for an AI-generated anticompetitive strategy?
• Is an algorithmic output attributable to the undertaking?
• Can autonomous learning constitute concerted practice?
• What compliance obligations should businesses impose on AI agents?
• Should agents be prevented from accessing certain competitor information?

These questions will become increasingly significant as AI systems become more autonomous.

23. Evidentiary Challenges

Signal-driven cases may require a new evidentiary toolkit.

Investigators may need:

Traditional evidence

• emails;
• contracts;
• meeting records;
• internal documents.

Technical evidence

• source code;
• API records;
• system logs;
• model documentation;
• training datasets;
• algorithm versions;
• audit trails.

Economic evidence

• pricing convergence;
• market response;
• deviation patterns;
• elasticity;
• profitability;
• structural breaks.

Behavioural evidence

• algorithm adoption;
• pricing responses;
• reaction times;
• deviations from independent pricing.

The combination of these forms of evidence may become more important than any single category.

24. Compliance Requirements for Businesses

Businesses operating in signal-driven markets should consider:

1. Information classification

Classify information as:

• public;
• historical;
• aggregated;
• commercially sensitive;
• strategically sensitive.

2. Algorithmic competition review

Before deploying pricing or recommendation algorithms, assess whether the system can facilitate coordination.

3. Third-party software due diligence

Review whether the vendor:

• collects competitors' data;
• aggregates market data;
• generates future pricing recommendations;
• uses customer data for model training.

4. Human oversight

Maintain meaningful human control over important competitive decisions.

5. Audit trails

Maintain records showing:

• input;
• recommendation;
• human decision;
• final outcome.

6. Competition-law training

Traditional cartel training should be expanded to cover:

• APIs;
• algorithms;
• data exchanges;
• automated pricing;
• AI agents;
• recommendation systems.

25. Regulatory Test for Signal-Driven Markets

A useful future analytical framework could be:

Step 1 — Identify the signal

What information is being generated?

Step 2 — Identify the controller

Who controls the signal?

Step 3 — Identify the recipients

Who receives it?

Step 4 — Assess sensitivity

Does it reveal competitively sensitive information?

Step 5 — Assess frequency

Is the information:

• historical;
• periodic;
• real-time;
• predictive?

Step 6 — Assess automation

Do firms merely observe the signal, or do algorithms automatically respond?

Step 7 — Assess market effects

Does the system:

• reduce uncertainty?
• facilitate coordination?
• foreclose rivals?
• increase switching costs?
• reinforce dominance?

Step 8 — Examine safeguards

Are there:

• firewalls?
• anonymisation?
• aggregation?
• human review?
• independent decision-making?

Step 9 — Select the legal theory

Potential theories include:

• cartel;
• concerted practice;
• information exchange;
• abuse of dominance;
• exclusionary conduct;
• tying;
• self-preferencing;
• refusal of access;
• merger-related data concentration.

26. Distinguishing Legitimate Signals from Anticompetitive Signals

Not every signal is harmful.

Potentially pro-competitive signals

• public prices;
• product quality information;
• safety information;
• genuine consumer reviews;
• publicly available market information;
• efficiency-enhancing demand forecasts.

These may reduce search costs and improve consumer welfare.

Potentially problematic signals

• confidential future pricing;
• confidential future output;
• competitor-specific strategic information;
• algorithmically coordinated pricing recommendations;
• signals deliberately designed to discourage competitive deviation.

Therefore, signal regulation should not simply prohibit transparency.

The objective should be to distinguish information that facilitates competition from information or systems that undermine independent competitive decision-making.

27. Key Competition-Law Principles Emerging

The case law and current enforcement developments suggest several principles:

1. Communication need not be human-to-human.
2. Digital platforms can transmit competitively significant signals.
3. Information exchange can be problematic even without an express price-fixing agreement.
4. Algorithms do not automatically immunise firms from cartel rules.
5. Control over rankings and recommendations can affect competition.
6. Dominant firms may possess power through information architecture rather than price alone.
7. Competitively sensitive data can become a mechanism of coordination.
8. Algorithmic feedback loops can amplify market power.
9. Auditability will become increasingly important.
10. Future regulation is likely to combine competition law with digital-market regulation.

28. Conclusion

Signal-driven markets represent a fundamental evolution in the economics of competition. Competitive behaviour is increasingly determined not merely by prices and quantities but by the information architecture surrounding firms and consumers.

The most important legal development is likely to be the movement from a narrow concept of “agreement” toward closer examination of:

information + algorithm + platform + market response.

The cases involving T-Mobile Netherlands, Eturas, Google Shopping, Airline Tariff Publishing and major digital-platform litigation, together with the ongoing RealPage algorithmic-pricing proceedings, provide important building blocks for this development. The RealPage proceedings are particularly significant because U.S. enforcement has directly addressed the alleged use of competitors' sensitive information through common pricing software.

The future challenge will be to preserve the benefits of real-time information, AI and automated decision-making while preventing those technologies from becoming mechanisms for price coordination, exclusion, discriminatory access or reinforcement of entrenched market power.