Competition Law and Hospitality Dynamic Pricing Algorithms

1. Introduction

Dynamic pricing has become central to the hospitality industry. Hotels, resorts, hotel chains, online travel agencies (OTAs), revenue-management companies and booking platforms increasingly use algorithms to determine room prices according to factors such as:

occupancy levels;

remaining room inventory;

booking pace;

seasonality;

competitor prices;

historical demand;

local events;

customer searches;

cancellation patterns;

length of stay;

distribution channel;

room type;

geographic demand; and

expected willingness to pay.

Dynamic pricing is not inherently anti-competitive. A hotel independently changing its price in response to genuine demand and supply is ordinarily legitimate.

The competition-law problem arises when the algorithm becomes a mechanism through which independent competitors coordinate prices, exchange competitively sensitive information, impose price parity, or reduce the incentive to compete.

The modern legal issue is therefore not simply:

“Did humans agree to fix prices?”

It is increasingly:

“Did competing businesses use technology in a manner that substitutes algorithmic coordination for independent price competition?”

This issue has become particularly important in hospitality. In Cornish-Adebiyi v. Caesars Entertainment, the U.S. DOJ and FTC specifically argued that hotels cannot use an algorithm to accomplish conduct that would be unlawful if performed by human employees. (Federal Trade Commission)

2. Meaning of Hospitality Dynamic Pricing Algorithms

A hospitality dynamic-pricing algorithm is a software system that continuously determines or recommends room rates.

Simplified model

Demand data + hotel inventory + market information + competitor information + historical data → Algorithm → Recommended room price

For example:

The first five examples generally represent ordinary competitive pricing.

The last two become legally sensitive when the algorithm uses non-public competitor information or systematically causes competitors to align their prices.

3. Competition-Law Framework

A. Horizontal price fixing

The most serious issue is an agreement between competing hotels to fix, maintain or coordinate prices.

Under U.S. law, this primarily implicates Section 1 of the Sherman Act.

Under EU law, the equivalent concern arises under Article 101 TFEU.

In India, comparable conduct can fall under Section 3 of the Competition Act, 2002, particularly agreements between enterprises that directly or indirectly determine purchase or sale prices.

The technological form of the agreement does not immunize it.

Thus:

Human price agreement

and

Algorithm-mediated price agreement

may produce the same competition-law problem.

4. Algorithmic Hub-and-Spoke Coordination

A particularly important model is the hub-and-spoke structure.

Structure

Hotel A
↓
Common pricing algorithm/provider
↑
Hotel B

and

Hotel C
↓
Same algorithm/provider

The algorithm provider functions as the hub, while competing hotels constitute the spokes.

The legal question is whether the spokes have effectively surrendered independent pricing decisions to a common mechanism.

The DOJ's RealPage case illustrates the concern: competing landlords supplied non-public pricing information to a common algorithmic system, which generated recommendations for participating competitors. The DOJ characterised this as resembling a classic hub-and-spoke arrangement. (Department of Justice)

The same economic logic can apply to hotel revenue-management software.

5. Six Major Case Laws

Case 1: Cornish-Adebiyi v. Caesars Entertainment

Importance

This is currently one of the most directly relevant cases for hotel algorithmic pricing.

The litigation concerned allegations that Las Vegas casino-hotels used Cendyn's Rainmaker revenue-management software and that the system generated room-price recommendations using information supplied by participating hotels.

The Third Circuit's 2026 decision recognised the significance of algorithmic coordination in the hotel sector. The allegations concerned hotels supplying current, non-public room-pricing and occupancy information to an AI-powered dynamic-pricing system which processed that information together with information from competitors. (Justia Law)

Legal principle

The important principle is that competitors cannot necessarily avoid Section 1 liability simply because:

they communicate through software;

the software provider calculates the prices;

hotels retain nominal discretion;

there is no traditional meeting or telephone call; or

the final price is technically entered by each hotel.

Competition-law significance

The case demonstrates the transition from traditional cartel law to algorithmic coordination theory.

For hospitality businesses, it means:

The use of a common pricing algorithm does not itself establish an unlawful cartel, but the surrounding agreements, information flows and conduct may establish one.

Case 2: Gibson v. Cendyn Group

The Gibson litigation concerns allegations that competing Las Vegas Strip hotels used Cendyn's revenue-management software and thereby contributed to inflated hotel room prices.

The allegations included a proposed hub-and-spoke conspiracy, with Cendyn functioning as the hub and participating hotels as spokes.

The Ninth Circuit's 2025 decision is particularly relevant because it examines the distinction between merely using common software and actually participating in an anticompetitive agreement. (Justia Law)

Key lesson

A common algorithm is not automatically unlawful.

A hotel may legitimately use the same commercially available software as competitors.

The competition problem becomes stronger where evidence shows:

competitors knowingly provide sensitive information;

the provider uses competitors' information to generate recommendations;

participants understand that competitors are doing the same;

participants agree to use the recommendations;

the system suppresses independent price competition; and

prices become coordinated as a result.

This distinction is extremely important.

Case 3: Eturas UAB and Others v. Lietuvos Respublikos konkurencijos taryba, C-74/14

This European Court of Justice case is extremely important for understanding technology-assisted concerted practices.

A group of travel agencies used a common computerised booking system. The system administrator sent a communication concerning the restriction of discounts, and the system automatically limited discounts available to customers. (EUR-Lex)

Principle

The ECJ considered whether participation in a common electronic system could constitute evidence of a concerted practice.

The case demonstrates that:

Competition law can attach legal significance to conduct occurring through a common digital infrastructure.

Hospitality relevance

Suppose several hotels use a common platform that automatically:

restricts discounts;

limits promotional prices;

adjusts rates;

synchronises pricing rules; or

communicates pricing restrictions.

The fact that the restriction is implemented automatically does not necessarily prevent competition-law scrutiny.

Key lesson

Automation does not eliminate the requirement of independent commercial decision-making.

Case 4: HRS – Hotel Reservation Service / Bundeskartellamt

The German HRS case concerned hotel price-parity or “best price” clauses.

HRS required hotels to provide it with the lowest room prices, maximum availability and favourable booking conditions available elsewhere.

The Düsseldorf Higher Regional Court confirmed the Bundeskartellamt's prohibition of HRS's best-price clauses. (Federal Cartel Office)

Competition concern

Such clauses can restrict competition between OTAs.

For example:

Hotel → Booking Platform A → ₹10,000
Hotel → Booking Platform B → ₹9,000

If the contract requires the hotel to offer Platform A the same lowest price, Platform B's ability to compete through lower prices is restricted.

Dynamic-pricing relevance

Algorithmic pricing can magnify this problem.

An OTA may observe:

hotel prices;

competitor prices;

availability;

booking velocity; and

demand.

A parity clause may then prevent the hotel from using its algorithm to offer differentiated prices through alternative channels.

Thus:

Dynamic pricing + parity restrictions = reduced pricing freedom.

Case 5: Booking.com – Bundeskartellamt

In 2015, the Bundeskartellamt prohibited Booking.com from continuing to use its “narrow” best-price clauses in Germany.

Under the narrow clause, hotels could offer lower prices on competing OTAs, but could not offer lower prices on their own websites than those displayed on Booking.com. (Federal Cartel Office)

Why this matters

The authority considered that even the narrow clause could:

restrict hotel pricing freedom;

reduce incentives for hotels to lower prices;

weaken competition between booking platforms; and

make entry by new platforms more difficult.

Dynamic algorithm connection

Imagine a hotel algorithm determining:

Direct website price = ₹8,500
OTA price = ₹9,000

If an OTA contract prohibits the hotel from offering ₹8,500 directly, the algorithm's ability to optimise across distribution channels becomes constrained.

This can reduce:

direct-booking competition;

OTA competition;

price differentiation;

consumer choice.

Case 6: Booking.com – France/Italy/Sweden Competition Authorities

The French, Italian and Swedish competition authorities, working with the European Commission, examined Booking.com's hotel price-parity practices.

Booking.com committed to modify its parity clauses and give hotels greater freedom to offer lower prices and better commercial conditions through competing channels. (Autorité de la Concurrence)

Significance

The authorities recognised that parity restrictions could:

reduce competition between OTAs;

hinder new platform entry;

restrict hotel pricing freedom; and

weaken competition over commissions and distribution.

Dynamic pricing significance

Dynamic-pricing systems depend upon the ability to react to different distribution channels.

A hotel may rationally wish to charge:

direct customer: ₹8,500;

OTA A: ₹9,000;

OTA B: ₹8,800;

corporate customer: ₹8,200;

loyalty member: ₹8,300.

Competition law may therefore protect the hotel's ability to differentiate its commercial strategy.

Case 7: United States v. RealPage

Although RealPage concerns residential rental housing rather than hotels, it is highly relevant to hospitality algorithmic pricing.

The DOJ alleged that competing landlords supplied RealPage with non-public, competitively sensitive information which was then used by algorithmic pricing software to generate recommendations. (Department of Justice)

The DOJ alleged violations of Sections 1 and 2 of the Sherman Act.

Hospitality application

Replace:

landlords → hotels

rent → room rate

apartment occupancy → hotel occupancy

The competition-law issue can be substantially similar.

A hotel algorithm becomes particularly problematic if:

Hotel A gives confidential occupancy and future-rate information to the algorithm provider → Hotel B does the same → algorithm aggregates information → algorithm recommends rates to both hotels.

That is considerably more problematic than an algorithm that merely uses publicly observable prices.

Case 8: District of Columbia v. RealPage

The District of Columbia litigation provides another useful illustration of algorithmic pricing theory.

The allegations concerned competitors delegating pricing decisions to a common pricing software system that analysed non-public information and generated pricing recommendations. (American Bar Association)

Relevance to hotels

The case highlights an important question:

How much independent pricing discretion must a hotel retain for algorithmic pricing to remain lawful?

Simply allowing a hotel manager to reject an algorithm's recommendation does not necessarily answer the question.

The broader issue is whether the commercial arrangement itself facilitates coordinated pricing.

6. Dynamic Pricing vs Algorithmic Collusion

It is essential to distinguish legitimate dynamic pricing from illegal coordination.

7. Why Competitor Data Is Particularly Dangerous

There is an important difference between public market intelligence and confidential competitor information.

Lower-risk information

A hotel may observe:

Hotel X publicly advertises ₹12,000.

It can independently decide to charge ₹11,500.

This is generally ordinary competitive behaviour.

Higher-risk information

A pricing provider receives:

Hotel X's future intended prices;

occupancy forecast;

unpublished discounts;

cancellation rates;

future inventory;

planned promotions;

confidential revenue data.

It then uses that information to recommend prices to Hotel Y.

The competition concern becomes much stronger.

8. Tacit Coordination

Algorithms may increase the possibility of tacit coordination.

Suppose:

Hotel A increases its price from ₹10,000 to ₹12,000.

Hotel B's algorithm immediately detects the change.

Hotel B follows.

Hotel A's algorithm then observes Hotel B's new price.

Hotel A follows again.

The result could be:

₹10,000 → ₹12,000 → ₹14,000 → ₹15,000

without a conventional cartel meeting.

But an important legal distinction must be maintained:

Parallel pricing alone is generally not sufficient to establish an unlawful agreement.

Competition authorities normally need evidence of communication, coordination, facilitating mechanisms, plus factors, or another legally sufficient basis for establishing an agreement or concerted practice.

That distinction is especially important in algorithm cases.

9. Price Signalling Through Algorithms

Another risk arises when algorithms make prices unusually transparent.

For example:

Hotel A publicly raises its price;

Hotel B's algorithm instantly observes it;

Hotel B raises its price;

Hotel A observes the response.

Rapid automated observation can make the market more predictable.

This can potentially make coordination easier because competitors no longer need to spend time discovering market behaviour.

The FTC has specifically warned that multiple sellers using the same pricing algorithm or providing competitively sensitive non-public information to a common pricing provider may create anticompetitive coordination concerns. (Federal Trade Commission)

10. Common Pricing Algorithms

The legal risk increases where many competing hotels use the same algorithm.

However:

Common software ≠ automatic cartel.

A common software product may legitimately provide:

forecasting;

occupancy optimisation;

inventory management;

demand prediction;

revenue forecasting.

The critical questions are:

What data does the software receive?

Is the data public or confidential?

Does the provider pool competitor information?

Does the algorithm recommend prices based on competitors' confidential data?

Are hotels encouraged to follow the recommendation?

Is there an agreement between competitors?

Can hotels independently reject recommendations?

Does the provider facilitate communication between competitors?

Does the system reduce independent price-setting?

Is the algorithm designed or marketed as a mechanism for industry-wide price optimisation?

11. Market Definition

Competition authorities must ordinarily identify the relevant market.

Possible markets include:

A. Hotel rooms

For example:

luxury hotel accommodation in a particular city.

B. Online hotel-booking services

The relevant market may include:

Booking platforms;

OTAs;

hotel websites;

other digital distribution channels.

C. Particular hotel segments

For example:

business hotels;

luxury hotels;

budget hotels;

resort accommodation.

Market definition matters because an algorithm may have different competitive effects depending upon market concentration.

12. Market Power

Algorithmic pricing is more concerning when the market is concentrated.

Suppose there are:

500 independent hotels

and each uses an independently configured algorithm.

Coordination may be more difficult.

Compare this with:

5 major hotel groups + one dominant pricing provider.

A common algorithm could potentially influence a large percentage of available room inventory.

The DOJ and FTC have expressly identified concentration and the ability of a small number of algorithm providers to influence substantial portions of a market as factors increasing algorithmic-collusion risk. (Federal Trade Commission)

13. Consumer Harm

Dynamic pricing can produce several forms of consumer harm.

1. Higher room prices

Coordinated algorithms may reduce price competition.

2. Reduced discounts

Hotels may stop competing aggressively through promotions.

3. Reduced output

Hotels may restrict available rooms at certain price points.

4. Reduced choice

Consumers may encounter similar prices across multiple hotels.

5. Reduced innovation

Smaller hotels or new OTAs may struggle to compete.

6. Increased intermediation costs

Parity clauses may prevent hotels from shifting customers toward cheaper direct channels.

14. Algorithmic Personalised Pricing

A separate issue is individualised or surveillance-based pricing.

An algorithm may use:

location;

browsing history;

loyalty status;

previous bookings;

device information;

customer profile;

search behaviour.

This creates a distinction between:

Dynamic pricing based on market demand

and

personalised pricing based on individual willingness to pay.

The latter can create additional consumer-protection, privacy and discrimination concerns, although those issues are not necessarily competition-law violations by themselves.

15. Indian Competition-Law Perspective

Under the Competition Act, 2002, hospitality dynamic pricing may be analysed principally through:

Section 3

Prohibits agreements that cause or are likely to cause an appreciable adverse effect on competition.

Particularly relevant are agreements involving:

price determination;

limitation of supply;

market allocation;

information exchange.

Section 4

Potentially becomes relevant where a dominant hotel group, OTA or digital intermediary abuses dominance.

Possible theories include:

discriminatory pricing;

unfair conditions;

exclusionary conduct;

denial of market access;

leveraging platform power.

Algorithmic pricing

An Indian competition authority could therefore examine:

Whether hotels are independently pricing rooms or whether a common technology provider has become a mechanism for coordinating prices.

16. Rule of Reason vs Per Se Approaches

Not every algorithmic-pricing case should automatically be treated as traditional price fixing.

Authorities may distinguish between:

Per se / hardcore coordination

Where competitors directly agree to fix prices.

Effects-based analysis

Where the alleged conduct involves:

common software;

information exchange;

vertical agreements;

parity clauses;

platform restrictions;

indirect coordination.

The U.S. RealPage litigation demonstrates the increasing importance of analysing algorithmic pricing through detailed evidence concerning information flows and competitive effects. (American Bar Association)

17. Evidence in Algorithmic Pricing Cases

Algorithm cases create unusual evidentiary questions.

Authorities may examine:

Software architecture

How does the algorithm actually work?

Data inputs

What information is uploaded?

Competitor information

Is information confidential?

Audit logs

Who accessed what information?

API records

Did competitors exchange information through the provider?

Contracts

What did the hotels agree with the software provider?

Internal communications

Did executives understand the competitive consequences?

Algorithm documentation

Was the system designed to coordinate or merely forecast?

Pricing outcomes

Did prices systematically converge after adoption?

18. Compliance Risks for Hotels

Hotels using dynamic pricing systems should adopt safeguards.

Recommended controls

Do not share non-public competitor pricing data unnecessarily.

Avoid agreements requiring competitors to follow common price recommendations.

Document independent pricing decisions.

Conduct algorithmic competition-law audits.

Separate public market information from confidential competitor information.

Review software-provider contracts.

Understand precisely what data the algorithm uses.

Avoid contractual restrictions that unnecessarily limit independent pricing.

Maintain human oversight.

Obtain competition-law review before deploying industry-wide pricing tools.

19. Compliance Checklist

Before implementing a hospitality dynamic-pricing system, ask:

Data

Is competitor data used?

Is it public?

Is it historical or current?

Is future pricing information collected?

Software

Is the same provider used by competitors?

Does the provider aggregate competitor data?

Does the provider know confidential pricing information?

Contracts

Are hotels required to use recommendations?

Are they required to maintain parity?

Are they prohibited from discounting independently?

Governance

Can management override the algorithm?

Are overrides recorded?

Is there an audit trail?

Has competition counsel reviewed the system?

Outcome

Are prices becoming unusually aligned?

Are discounts disappearing?

Are competitors responding almost instantaneously to each other's price changes?

20. Key Legal Principles Emerging from the Cases

The cases collectively establish several important propositions.

Principle 1 — Technology is not a defence

A business cannot escape competition law merely because an algorithm performs the coordination.

Principle 2 — Common software is not automatically illegal

The use of the same software by competitors is not, by itself, proof of a cartel.

Principle 3 — Confidential competitor information is highly sensitive

The aggregation of non-public competitor pricing information creates substantial antitrust risk.

Principle 4 — Human communication is not essential in every case

Digital systems can provide evidence of a concerted practice or facilitate coordination.

Principle 5 — Price-parity clauses can restrict competition

The HRS and Booking.com proceedings demonstrate that contractual restrictions can limit both OTA competition and hotel pricing freedom. (Federal Cartel Office)

Principle 6 — Independent pricing must remain genuine

A hotel should genuinely retain the ability and incentive to make its own pricing decision.

Principle 7 — Algorithmic neutrality matters

A system designed to optimise each hotel's independent position is less problematic than a system designed around the collective optimisation of participating competitors.

21. A Useful Hypothetical

Assume five hotels in Delhi use the same revenue-management company.

The provider receives:

daily occupancy;

future room inventory;

future intended prices;

cancellation rates;

promotional plans.

The provider's algorithm recommends:

“All participating hotels should increase the price by 15% for the upcoming festival.”

If the hotels knowingly participate in this system and rely upon competitor information to establish their prices, authorities could investigate a possible hub-and-spoke pricing arrangement.

Now change the facts.

Each hotel separately provides only:

its own occupancy;

its own historical bookings;

publicly available market data.

The software generates independent recommendations for each hotel without sharing confidential information between competitors.

The competition-law risk is substantially lower.

22. Important Distinction: Dynamic Pricing Is Not Price Fixing

This is perhaps the most important conclusion.

Legitimate

“Demand is high, so our algorithm increases our room price.”

Potentially problematic

“Our algorithm knows our competitors' confidential future prices and recommends a common price to participating hotels.”

The first is ordinary competitive conduct.

The second potentially raises serious cartel and information-exchange concerns.

23. Conclusion

Hospitality dynamic pricing algorithms represent a significant development in competition law because they can simultaneously increase pricing efficiency and facilitate coordination.

The law does not prohibit algorithms merely because they produce higher or changing prices. The critical issue is whether the technology preserves independent competitive decision-making.

The most important lessons from Cornish-Adebiyi, Gibson, Eturas, HRS, Booking.com and RealPage are that:

algorithmic conduct is subject to ordinary competition law;

common pricing software does not automatically create liability;

non-public competitor data is particularly dangerous;

hub-and-spoke algorithmic coordination can attract cartel scrutiny;

price-parity restrictions can reduce hotel and platform competition;

retaining nominal discretion may not eliminate liability where the overall arrangement coordinates pricing;

algorithm design, contracts, data inputs and actual use must be examined together; and

hotels should be able to demonstrate genuine, independent pricing decisions.

Accordingly, the central competition-law test for hospitality algorithms can be expressed as:

Independent data + independent algorithmic decision + independent pricing = generally lower competition risk.

Whereas:

Competitor confidential data + common pricing mechanism + coordinated reliance = substantially higher antitrust risk.

The current U.S. hotel litigation makes this particularly clear: regulators have expressly stated that hotels cannot use algorithms to accomplish conduct that would be unlawful if c