Competition Law and Governance of Observation-Driven Competition

1. Introduction

Observation-driven competition describes markets in which firms compete by continuously observing, collecting, analysing and reacting to information about competitors, customers and market conditions.

Traditional competition assumes that firms make largely independent decisions with imperfect information. Digital and data-intensive markets can change that structure. Firms may observe:

• competitors’ prices and discounts;
• inventory and capacity;
• customer searches and clicks;
• bids and offers;
• product rankings;
• demand forecasts;
• employee compensation;
• transaction histories;
• algorithmic responses of rivals;
• real-time market conditions.

Observation itself is not unlawful. Competition law becomes concerned where observation reduces strategic uncertainty so substantially that competitors can coordinate rather than compete independently, or where a dominant platform uses observational data to exclude rivals.

The legal problem can therefore be expressed as:

Observation → information advantage → prediction → strategic reaction → possible coordination or exclusion.

The central governance question is how competition law can preserve the benefits of information and data while preventing observation from becoming a mechanism for collusion, exclusion or exploitation.

2. Meaning of Observation-Driven Competition

Observation-driven competition exists where competitive decisions are materially shaped by systematic observation of market behaviour.

Traditional competition

A firm may ask:

"What price should we charge?"

Observation-driven competition

The firm may ask:

"What price is my competitor charging right now, how are customers responding, what inventory does the competitor have, and what price will its algorithm probably choose next?"

The difference is important because modern technology can make market observation:

1. continuous rather than occasional;
2. automated rather than human;
3. real-time rather than historical;
4. individualised rather than aggregated;
5. predictive rather than merely descriptive; and
6. self-reinforcing, because observations are fed back into algorithms.

3. Legal Framework

Observation-driven competition is primarily governed through established competition-law doctrines rather than an entirely separate body of law.

A. Horizontal agreements and concerted practices

Information exchanged between competitors can facilitate:

• price fixing;
• output restriction;
• market allocation;
• bid coordination;
• customer allocation;
• wage coordination;
• coordinated reduction of discounts.

Under EU law, Article 101 TFEU addresses agreements, decisions and concerted practices.

Under U.S. law, Section 1 of the Sherman Act addresses agreements restraining trade.

The important issue is not simply whether data were exchanged, but:

• what information was exchanged;
• whether it was competitively sensitive;
• whether it was current or historical;
• whether it identified individual competitors or customers;
• how frequently it was exchanged;
• whether the market was concentrated;
• whether firms used the information to coordinate behaviour.

4. Information Transparency and Competitive Uncertainty

Competition does not necessarily require complete secrecy.

Some transparency can benefit consumers because it enables:

• price comparison;
• product comparison;
• informed purchasing;
• lower search costs;
• improved market entry.

But excessive transparency between competitors can have the opposite effect.

Competitive uncertainty

In a competitive market, Firm A may not know:

"Will Firm B reduce its price tomorrow?"

That uncertainty can force Firm A to compete aggressively.

If an information system allows Firm A to observe Firm B's price instantaneously, Firm A can react immediately.

If all firms have access to the same information, the market may become highly transparent to competitors.

This can make tacit or explicit coordination easier.

5. Case Law

1. United States v. Container Corporation of America, 393 U.S. 333 (1969)

This is a foundational U.S. case concerning competitor-to-competitor exchange of price information.

Competitors in the corrugated-container industry exchanged information concerning prices charged to particular customers. The Supreme Court held that the exchange constituted concerted action and could have an anticompetitive effect because knowledge of competitors' prices reduced the intensity of price competition.

Principle

Information exchange can itself become a competition concern where it facilitates coordinated pricing.

Relevance to observation-driven competition

Modern platforms can perform electronically what firms historically did manually:

Observe → communicate → compare → react.

The digitalisation of the observation mechanism does not automatically remove it from antitrust scrutiny.

2. Todd v. Exxon Corp., 275 F.3d 191 (2d Cir. 2001)

In Todd v. Exxon, employees alleged that major oil and petrochemical companies exchanged detailed information concerning compensation for managerial, professional and technical employees.

The Second Circuit held that the complaint adequately alleged a Sherman Act §1 information-exchange claim because it identified a plausible market susceptible to collusion, information exchange with anticompetitive potential and resulting antitrust injury.

Principle

Information exchange is not limited to product prices.

It may concern:

• wages;
• salaries;
• employment terms;
• costs;
• customers;
• production;
• strategic variables.

Observation-driven significance

Observation-driven competition therefore extends to labour markets.

A digital labour-market platform that systematically provides competing employers with highly detailed current compensation information may potentially alter competitive incentives in the same way that product-price information can.

3. United States v. Airline Tariff Publishing Co. et al., D.D.C. 1993–1994

The Airline Tariff Publishing Company system allowed airlines to disseminate fare information throughout the industry.

The U.S. Department of Justice brought an antitrust action alleging price-fixing involving major airlines and the tariff-publishing system. The case resulted in a final judgment restricting specified conduct.

Principle

A common information infrastructure can become problematic when it allows competitors to communicate or signal competitively significant pricing information in a manner that facilitates coordination.

Observation-driven significance

This case is particularly important because it resembles a modern digital marketplace:

Common information infrastructure → real-time observation → strategic response.

Modern platforms may perform this function at dramatically greater speed and scale.

4. T-Mobile Netherlands BV and Others v. Netherlands Competition Authority, C-8/08 (2009)

The Court of Justice of the European Union considered the concept of a concerted practice under Article 81 EC, now Article 101 TFEU.

The case concerned communications among mobile telecommunications operators and the question whether a single meeting could constitute sufficient coordination. The Court recognised that a single contact can, depending upon its content and purpose, be sufficient to establish a concerted practice where the legal conditions are satisfied.

Principle

Competition law does not necessarily require a long-running formal agreement.

Observation-driven significance

This matters in automated markets because coordination can occur rapidly.

A platform or algorithm does not necessarily need years of communications to produce coordinated conduct. A short exchange of strategically sensitive information can potentially alter firms' subsequent behaviour.

5. Eturas UAB and Others v. Lithuanian Competition Council, C-74/14 (2016)

This is one of the most important European cases for technology-mediated coordination.

Travel agencies used a common computerised booking system. The system administrator sent a message concerning restrictions on discounts, and the system automatically restricted the discounts available to customers. The Court considered whether this could constitute a concerted practice and addressed the evidentiary requirements for attributing knowledge and participation to individual undertakings.

Principle

The technological architecture through which firms interact can become relevant evidence of coordination.

Importantly, the case also demonstrates that mere participation in a technological system is not automatically equivalent to participation in an unlawful agreement. Evidence concerning awareness and participation remains important.

Observation-driven significance

The case provides a useful framework for modern platforms:

common software + communication + automated implementation + competitor participation

may create competition-law questions even without a conventional face-to-face cartel meeting.

6. Google and Alphabet v. European Commission — Google Shopping, T-612/17

In Google Shopping, the EU General Court examined Google's treatment of competing comparison-shopping services within its general search service.

The Court upheld the finding that Google had abused its dominant position by favouring its own specialised search service in the display of results, rather than treating competing services on equivalent terms.

Principle

A dominant undertaking can infringe competition law through the way it operates an important digital infrastructure, particularly where its control over access or ranking advantages its own service.

Observation-driven significance

Search platforms observe enormous quantities of:

• queries;
• clicks;
• rankings;
• user behaviour;
• product information;
• competitor performance.

That observational capacity can become a competitive advantage.

The competition issue therefore shifts from simple data collection to:

Who observes the market, who controls the resulting information, and how that information is used to compete?

7. Gibson v. Cendyn Group, LLC, 148 F.4th 1069 (9th Cir. 2025)

Gibson v. Cendyn is particularly relevant to modern algorithmic pricing.

The litigation concerned allegations involving hotel pricing algorithms and the use of competitor information. The Ninth Circuit's 2025 decision became an important appellate development in U.S. algorithmic-pricing litigation. Contemporary antitrust analysis identifies it as the first U.S. appellate decision specifically addressing algorithmic-pricing litigation.

Principle

The use of an algorithm does not itself determine whether conduct is lawful or unlawful.

The important questions include:

• what information enters the algorithm;
• who supplies that information;
• whether competitors' sensitive information is used;
• whether the system facilitates coordination;
• whether an agreement or concerted practice can be established.

Observation-driven significance

This illustrates the movement from:

human observation → human pricing decision

to:

automated observation → algorithmic prediction → algorithmic pricing.

8. United States and States v. RealPage, Inc.

Although this matter is primarily an enforcement proceeding rather than a traditional reported judicial precedent, it is highly relevant to the contemporary governance of observation-driven competition.

The U.S. Department of Justice alleged that RealPage's pricing software collected non-public, competitively sensitive information from competing landlords and used that information to generate rental-price recommendations. The government alleged violations of Sections 1 and 2 of the Sherman Act.

The subsequent enforcement process produced settlements and judgments involving participating landlords and RealPage. By September 2026, DOJ case materials included judgments and proposed settlements addressing the use of competitors' sensitive information and algorithmic pricing.

Principle

Competition law can apply where:

competitors supply sensitive information to a common algorithm that then influences their competitive decisions.

The important point is that the algorithm is not treated as a legal shield.

Observation-driven significance

RealPage illustrates the full observation loop:

Competitor data → central database → algorithmic analysis → price recommendation → competitor response → new market data.

That feedback loop can potentially make coordination more persistent and more difficult for individual firms to resist.

6. Major Competition Concerns

A. Algorithmic collusion

Algorithms can observe prices continuously and react almost instantaneously.

Potential mechanism:

Firm A raises price → algorithm observes → Firm B's algorithm responds → Firm A observes → repeated price alignment.

The legal difficulty is distinguishing:

1. independent algorithmic optimisation;
2. conscious parallelism;
3. algorithm-assisted tacit coordination; and
4. an actual agreement or concerted practice.

B. Exchange of competitively sensitive information

Particular risks arise from information concerning:

• current prices;
• future prices;
• discounts;
• customer-specific terms;
• capacity;
• inventories;
• costs;
• strategic plans;
• bids;
• wages.

The more current, granular, individualised and strategically sensitive the information, the greater the potential competition concern.

C. Common algorithm providers

A single technology provider serving many competitors can create a structural issue.

For example:

100 competing landlords → one pricing algorithm → common information pool → recommendations to all landlords.

The technology provider may effectively become an information intermediary connecting competitors.

D. Self-learning algorithms

Machine-learning systems create an additional issue.

An algorithm may learn from:

• previous competitor prices;
• consumer reactions;
• supply conditions;
• historical discounts;
• competitors' responses.

Consequently, the algorithm can develop sophisticated predictions without a human explicitly instructing it to coordinate.

Competition law must therefore examine system design and data architecture, not merely written instructions.

7. Observation and Dominance

Observation-driven competition also creates Article 102 / Section 2-type concerns where a dominant undertaking controls an important information infrastructure.

Potential theories include:

1. Self-preferencing

A platform uses information obtained from third-party participants to improve its own competing service.

2. Data advantage

A dominant firm obtains market intelligence unavailable to competitors.

3. Discriminatory access

The platform gives itself or selected firms superior access to data, APIs or analytical tools.

4. Leveraging

Information collected in one market is used to strengthen dominance in another.

5. Exclusion

Competitors become unable to compete because they cannot replicate the dominant firm's observational dataset.

8. Observation as a Competitive Asset

Modern competition increasingly involves competition over observational capacity.

A firm may possess:

Thus, competition may increasingly occur not merely over products but over:

the ability to observe the market itself.

9. Governance of Observation-Driven Competition

Effective governance should operate at several levels.

A. Data governance

Companies should classify information according to competitive sensitivity.

Low-risk

• old aggregated statistics;
• publicly available information;
• broad industry trends.

Higher-risk

• current individual prices;
• future pricing plans;
• customer-specific information;
• individual competitor inventory;
• future capacity decisions.

B. Algorithm governance

Businesses should document:

• data inputs;
• data sources;
• model objectives;
• optimisation parameters;
• competitor-data use;
• automated pricing rules;
• human override mechanisms.

C. Information-exchange controls

Competitors should generally avoid unnecessary exchange of:

• future prices;
• strategic plans;
• individual customer information;
• current costs;
• capacity plans;
• commercially sensitive forecasts.

Where legitimate industry data sharing exists, safeguards may include:

• aggregation;
• anonymisation;
• historical data;
• independent intermediaries;
• access controls.

10. Competition Compliance for AI and Algorithms

A useful compliance framework is:

Step 1 — Identify the observation

What exactly does the system observe?

Step 2 — Identify the source

Is the information:

• public;
• privately obtained;
• competitor-supplied;
• customer-generated?

Step 3 — Determine sensitivity

Is it strategically significant?

Step 4 — Examine the algorithm

Does the algorithm merely analyse information, or does it facilitate coordinated conduct?

Step 5 — Examine the market

Consider:

• concentration;
• barriers to entry;
• frequency of interaction;
• transparency;
• switching costs;
• number of competitors.

Step 6 — Examine feedback loops

Does the algorithm continuously observe competitors and adjust decisions?

Step 7 — Establish human accountability

Who designed, approved and supervised the system?

11. Key Doctrinal Distinction

The most important distinction is:

Lawful observation

"We monitor publicly available market prices so that we can compete more effectively."

versus

Potentially unlawful coordination

"We obtain competitors' confidential pricing information through a common system and use it to align our pricing decisions."

The technological sophistication of the mechanism does not by itself change the underlying competition-law analysis.

12. Emerging Issues

Observation-driven competition raises several unresolved or developing questions.

1. Autonomous coordination

Can an algorithm's independent learning create legally significant coordination without an explicit human agreement?

2. Predictive competitor intelligence

When does prediction of competitors become equivalent to obtaining competitively sensitive information?

3. Data pooling

When does a shared industry database become a legitimate efficiency-enhancing facility and when does it become a coordination mechanism?

4. Platform neutrality

Can a marketplace use information generated by sellers to compete against those same sellers?

5. Real-time markets

Does extreme transparency reduce competitive uncertainty to an anticompetitive degree?

6. AI governance

Should competition compliance become part of the design stage of AI systems rather than merely an after-the-fact legal review?

13. Relationship Between the Major Cases

The cases can be understood as a developing legal chain:

Container Corp.
→ competitor price information can reduce competitive intensity

Todd v. Exxon
→ information exchange can extend beyond product prices

Airline Tariff Publishing
→ common information infrastructure can facilitate coordination

T-Mobile Netherlands
→ concerted practices need not involve prolonged coordination

Eturas
→ automated technological systems can form part of the coordination mechanism

Google Shopping
→ control over digital information infrastructure can generate dominance concerns

Gibson v. Cendyn
→ algorithmic pricing brings these traditional principles into automated markets

RealPage
→ current enforcement increasingly addresses the combination of sensitive competitor data and algorithmic pricing

14. Conclusion

Observation-driven competition is not inherently anticompetitive. Market observation can produce substantial efficiencies, improve price comparison, reduce search costs and enable firms to respond rapidly to consumer demand.

The competition-law concern arises when observation changes the structure of competitive decision-making.

The critical risks are:

1. exchange of competitively sensitive information;
2. algorithmic coordination;
3. reduction of strategic uncertainty;
4. common algorithmic intermediaries;
5. self-preferencing based on privileged observation;
6. data-driven exclusion;
7. feedback loops between competitors; and
8. use of dominant information infrastructures to disadvantage rivals.

The central legal principle emerging from the case law is that competition law follows the economic function of information, not merely the technological form in which that information is processed. A spreadsheet, API, marketplace, pricing engine or AI model can all raise essentially the same competition-law question: does the information mechanism preserve independent competitive decision-making, or does it facilitate coordination or exclusion?