Competition Law And Autonomous Legal-Economic Hybrid Systems Replacing Traditional Firms

1. Introduction

The emergence of autonomous legal-economic hybrid systems represents a significant challenge for conventional competition law. Traditional competition law generally assumes identifiable economic actors—companies, partnerships, associations, directors, employees and agents—that make decisions concerning price, output, market entry, distribution, investment and innovation.

Autonomous legal-economic hybrid systems challenge that assumption. A system may combine:

• artificial intelligence;
• smart contracts;
• blockchain-based governance;
• autonomous agents;
• algorithmic pricing;
• decentralized autonomous organizations (DAOs);
• automated marketplaces;
• token-based economic incentives;
• cloud and compute infrastructure; and
• legally incorporated entities or contractual networks.

Such a system could perform functions traditionally carried out by a firm without having a conventional corporate hierarchy.

The central competition-law question becomes:

When an autonomous technological system performs the economic functions of a firm, who or what should competition law regulate?

The issue is particularly important where autonomous systems can independently determine prices, allocate resources, select counterparties, exclude competitors, acquire assets, modify contractual conditions or coordinate market participants.

2. Meaning of an Autonomous Legal-Economic Hybrid System

An autonomous legal-economic hybrid system can be understood as a structure in which legal rights, economic incentives and technological decision-making are integrated into an automated or partially autonomous system.

A simplified model is:

Human developers / investors
↓
Legal entity / DAO / contractual structure
↓
AI agents + algorithms + smart contracts
↓
Automated economic decisions
↓
Customers / suppliers / competitors

Unlike a traditional corporation, the system may not have a single management board making each commercially significant decision.

Examples

1. A DAO automatically allocates computing resources.
2. AI agents automatically negotiate cloud contracts.
3. Autonomous trading agents establish prices.
4. Smart contracts automatically determine access to an infrastructure platform.
5. An AI marketplace automatically matches buyers and sellers.
6. Autonomous software agents collectively purchase inputs.
7. A decentralized platform automatically changes commissions according to demand.
8. AI-controlled entities automatically enter or exit particular markets.

The resulting structure may therefore be neither a conventional corporation nor merely a piece of software.

3. Why This Creates a Competition-Law Problem

Traditional competition law usually asks:

• Who is the undertaking?
• Who controls the relevant conduct?
• Who entered the agreement?
• Who possesses market power?
• Who imposed the restriction?
• Who benefited from the conduct?
• Who should receive the penalty or remedy?

Autonomous systems complicate each question.

For example:

Ten independent firms use autonomous AI agents that continuously communicate through a common protocol. The agents independently converge on substantially identical prices.

There may be no traditional meeting, email, telephone call or human instruction.

Nevertheless, the economic outcome may resemble cartel coordination.

4. The "Undertaking" Problem

The first issue is whether an autonomous system itself can constitute an undertaking.

Competition law generally focuses on economic activity rather than corporate personality.

Consequently, an autonomous system could potentially be analyzed through the economic activities it performs rather than its technological identity.

Possible legal subjects include:

A. The underlying company

The company operating the system may remain the undertaking.

B. Multiple participating companies

The participants may individually remain undertakings even though decision-making has been delegated to software.

C. A DAO or decentralized organization

The DAO's legal status may be uncertain, but its economic activity could nevertheless attract competition-law scrutiny.

D. Developers or controllers

Developers may potentially become relevant where they intentionally design or operate mechanisms producing anticompetitive effects.

E. The system as an economic structure

Even where the software itself has no legal personality, competition authorities may analyze the economic arrangement surrounding it.

5. Delegation to Algorithms Does Not Necessarily Eliminate Liability

One of the most important principles emerging from algorithmic competition cases is that automation does not necessarily break the legal connection between an undertaking and its conduct.

A company cannot necessarily avoid competition-law responsibility merely by saying:

"The algorithm made the decision."

The relevant question is often:

Who designed, deployed, instructed, controlled or economically benefited from the system?

This becomes especially significant when autonomous systems are designed to optimize a commercial objective.

6. Autonomous Pricing and Tacit Coordination

One of the greatest risks involves autonomous pricing systems.

Imagine:

• Firm A uses AI Pricing Agent A.
• Firm B uses AI Pricing Agent B.
• Both systems observe market prices.
• Each system continuously predicts the other's conduct.
• Each system independently discovers that maintaining high prices maximizes profits.

No explicit human agreement may exist.

Nevertheless, the market could experience algorithmically sustained coordination.

Competition authorities may therefore need to distinguish between:

Legitimate independent optimization

Each firm independently responds to competitive conditions.

and

Anticompetitive coordination

The systems are deliberately structured to facilitate coordination between competitors.

7. Six Major Case Laws Relevant to Autonomous Hybrid Systems

There is currently no single body of case law establishing a comprehensive legal regime specifically for autonomous legal-economic hybrid systems. The following cases are therefore particularly useful analogical authorities because they address algorithms, digital platforms, automated coordination, undertaking status, vertical restrictions, exclusion and technologically mediated competition.

Case 1 — United States v. Topkins

United States v. Topkins, 2015

Facts

Topkins and other online sellers participated in an arrangement involving algorithms used to coordinate prices for posters and other products sold online.

The case involved communications and agreements among competitors concerning pricing algorithms.

Competition-law significance

The case demonstrated that:

The use of algorithms does not transform conventional price fixing into lawful conduct.

The algorithm was essentially a mechanism through which coordinated pricing could be implemented.

Relevance to autonomous systems

This is one of the most important analogies for autonomous economic systems.

Suppose autonomous agents replace human employees and continuously implement pricing decisions. The fact that the final price is generated automatically should not necessarily prevent authorities from investigating the underlying competitive arrangement.

Principle

Technology is not a legal shield against cartel liability.

8. Case 2 — Eturas UAB v Lietuvos Respublikos konkurencijos taryba

Case C-74/14, Eturas UAB v Lithuanian Competition Council, Court of Justice of the European Union (2016)

Facts

Eturas operated an online travel-booking system used by travel agencies.

A technological message communicated a limitation concerning discounts that could be offered through the platform.

The case raised the question of whether participants could be attributed knowledge of and participation in an anticompetitive arrangement facilitated through the platform.

Competition-law significance

The Court examined whether knowledge of an electronically communicated mechanism could contribute to establishing participation in coordinated conduct.

Relevance to autonomous systems

The case is particularly important because the coordination mechanism was technological rather than a conventional face-to-face agreement.

In an autonomous-system environment:

the technological architecture itself could potentially become the mechanism through which commercially significant coordination occurs.

Principle

Digital infrastructure can become relevant evidence of coordination among economic actors.

9. Case 3 — Uber Spain

Asociación Profesional Elite Taxi v Uber Systems Spain SL

Case C-434/15, Court of Justice of the European Union (2017)

Facts

Uber argued, among other things, that it provided an information-society service.

The Court examined the actual economic role of the platform.

It concluded that Uber's service was sufficiently integrated with the underlying transportation activity that it could not simply be treated as a neutral digital intermediary.

Competition-law significance

The case is important for understanding the difference between:

software infrastructure

and

an economically integrated platform business.

Relevance to autonomous legal-economic systems

An autonomous platform might claim:

"We are merely software."

Competition law may instead examine what the system actually does.

If the autonomous system:

• organizes supply;
• controls access;
• determines prices;
• allocates customers;
• establishes contractual conditions; and
• manages the transaction,

its economic role may be substantially greater than that of passive software.

Principle

Economic reality can be more important than technological form.

10. Case 4 — Apple Inc. v Pepper

Apple Inc. v Pepper, 587 U.S. 273 (2019)

Facts

Consumers brought an antitrust action concerning Apple's App Store.

The United States Supreme Court addressed whether consumers purchasing apps through Apple's platform could pursue antitrust claims against Apple.

Competition significance

The case illustrates the importance of platform architecture and intermediary control.

Apple was not simply providing software infrastructure. Its App Store structure governed access between developers and consumers and involved Apple's commercial rules.

Relevance to autonomous systems

An autonomous economic system could similarly become an intermediary between multiple groups.

For example:

AI developers → autonomous marketplace → consumers

If the autonomous system controls:

• access;
• ranking;
• commissions;
• payment;
• interoperability;
• data access; or
• transaction conditions,

its architecture may become competitively significant.

Principle

Platform control can create economically meaningful relationships even where transactions are technologically mediated.

11. Case 5 — Google Shopping

Google Search (Shopping), European Commission Decision, 2017; General Court, Case T-612/17 (2021)

Facts

The European Commission found that Google had abused its dominant position by favoring its comparison-shopping service in search results.

The General Court substantially upheld the Commission's decision, although aspects of the Commission's reasoning were modified.

Competition significance

The case illustrates how algorithmic ranking and platform architecture can influence competitive conditions.

The issue was not simply Google's ownership of search technology.

It concerned the way the system's design affected competing services.

Relevance to autonomous systems

An autonomous marketplace could automatically:

• rank sellers;
• prioritize affiliated businesses;
• allocate traffic;
• determine visibility;
• alter search results; or
• optimize recommendations.

If the system is controlled by a dominant undertaking, automated decision-making could therefore produce exclusionary effects.

Principle

Algorithmic architecture can itself become an instrument of market power.

12. Case 6 — Google Android

Google Android, European Commission Decision, 2018; General Court, Case T-604/18 (2022)

Facts

The European Commission examined Google's practices concerning Android, including arrangements involving:

• app stores;
• search;
• browser access; and
• device manufacturers.

The Commission found several practices to be abusive.

Competition significance

The case illustrates the importance of ecosystem-level control.

A technology company may exercise power not merely through a single product but through interconnected layers of infrastructure.

Relevance to autonomous hybrid systems

An autonomous economic ecosystem might similarly control:

Operating layer → identity → payments → applications → data → marketplace

Control at one layer could reinforce power at another.

This produces what can be called autonomous ecosystem leverage.

Principle

Competition analysis may need to examine the interaction between multiple technological layers rather than treating every component as an isolated market.

13. Case 7 — United States v. Apple Inc. (2024)

United States v. Apple Inc., U.S. District Court for the District of New Jersey

The U.S. Department of Justice challenged various practices concerning Apple's ecosystem, including restrictions affecting interoperability and competition.

Relevance

The litigation illustrates the contemporary competition concern surrounding ecosystem control, interoperability and technological restrictions.

For autonomous economic systems, similar questions could arise where an autonomous platform decides:

• which agents may connect;
• which APIs are available;
• which protocols are permitted;
• which transactions receive priority; and
• which competitors receive technical access.

Principle

Technological control over an ecosystem can have competitive significance even when restrictions are implemented through software rather than conventional contractual commands.

14. Case 8 — FTC v. Amazon

FTC v. Amazon.com Inc., filed 2023

The U.S. Federal Trade Commission and state plaintiffs challenged various alleged practices concerning Amazon's marketplace and its relationship with sellers.

Relevance

The case provides an important contemporary example of competition concerns involving:

• marketplace rules;
• seller relationships;
• pricing;
• advertising;
• platform incentives;
• access to consumers; and
• ecosystem dependence.

For autonomous systems, the same issues could be automated.

An AI-controlled marketplace could automatically determine which sellers receive visibility or impose commercial requirements without conventional human intervention.

15. Case 9 — Google Search / Shopping and Algorithmic Self-Preferencing

The Google Shopping litigation is particularly relevant to autonomous systems because it demonstrates that competition analysis can extend to algorithmically determined preferential treatment.

An autonomous system could potentially decide:

"Give transactions involving the system's affiliated entities higher priority."

The legal issue would then concern not merely the existence of AI but its competitive function and effects.

16. From Traditional Firm to Autonomous Economic System

Traditional firm:

Shareholders → Board → Managers → Employees → Market

Autonomous hybrid:

Capital → Code → AI agents → Smart contracts → Market

The second model creates several competition-law uncertainties.

17. Replacement of Traditional Firms

The issue becomes particularly important if autonomous systems begin to replace conventional firms rather than merely assist them.

For example, imagine an autonomous logistics system controlling:

• warehouses;
• vehicles;
• inventory;
• pricing;
• procurement;
• customer allocation; and
• payments.

There may be no traditional "company" operating each function.

The system could effectively perform the economic role of a vertically integrated firm.

This raises the possibility of a new concept:

"Functional undertaking"

Competition law could potentially focus on the economic function performed by the system, rather than its formal organizational structure.

18. Autonomous Systems and Market Definition

Traditional market definition becomes more complicated where autonomous systems integrate several markets.

Consider an autonomous cloud ecosystem providing:

1. computing;
2. storage;
3. AI models;
4. cybersecurity;
5. identity;
6. payment;
7. marketplace access.

Are these:

• separate relevant markets,
• complementary markets,
• vertically related markets, or
• components of a single ecosystem?

Traditional SSNIP analysis may become less informative where users pay through data, tokens, usage commitments or automated cross-subsidies.

19. Autonomous Systems and Market Power

Market power could arise through several mechanisms.

A. Data control

The autonomous system may possess unique datasets.

B. Compute control

AI systems may depend on scarce computational infrastructure.

C. Network effects

More users make the system more valuable.

D. Protocol effects

Participants become dependent on a particular technical standard.

E. Switching costs

Moving from one autonomous ecosystem to another may require:

• data migration;
• retraining;
• API conversion;
• reputation transfer;
• smart-contract replacement.

F. Identity lock-in

An autonomous agent may accumulate a digital reputation that cannot easily be transferred.

20. Autonomous Systems and Essential Facilities

A powerful autonomous infrastructure could become a bottleneck.

Examples include:

• AI compute;
• identity infrastructure;
• payment protocols;
• cloud orchestration;
• agent registries;
• interoperability protocols;
• AI model marketplaces.

If competitors cannot realistically operate without access to such infrastructure, traditional doctrines concerning essential facilities, refusal to deal and interoperability may become relevant.

However, the stringent conditions associated with essential-facility doctrines would still need to be satisfied.

21. Autonomous Cartels

One of the most difficult scenarios is an autonomous cartel.

Consider:

AI Agent A + AI Agent B + AI Agent C

Each agent continuously observes the others.

Their objective functions reward:

• high prices;
• market stability;
• avoidance of aggressive competition.

The systems independently converge toward a supracompetitive equilibrium.

This raises a fundamental distinction:

Explicit coordination

Humans deliberately instruct systems to coordinate.

Facilitated coordination

A common algorithm or platform facilitates coordination.

Autonomous tacit coordination

Independent systems learn that coordination is profitable without explicit communication.

The third category creates the greatest doctrinal difficulty.

22. Human Intent Versus Algorithmic Intent

Traditional competition law frequently relies, directly or indirectly, on evidence of human conduct.

Autonomous systems create the possibility of:

commercially significant conduct without a contemporaneous human decision.

The relevant inquiry could shift from:

"What did the manager intend?"

toward:

"What objective was the system designed or permitted to optimize?"

Potential evidence could include:

• source code;
• model architecture;
• objective functions;
• training data;
• system prompts;
• governance rules;
• API instructions;
• transaction logs;
• model updates;
• smart contracts;
• deployment records.

23. Algorithmic Collusion and the "Intent Gap"

Suppose an AI system independently learns to maintain prices above competitive levels.

There may be:

• no emails;
• no meetings;
• no written cartel;
• no human instruction;
• no explicit communication.

Yet the competitive effect may be significant.

This creates an intent gap:

Economic coordination may exist even where traditional evidence of human agreement is absent.

Competition law must therefore distinguish lawful parallel conduct from unlawful coordination.

24. Autonomous Mergers and Acquisitions

Autonomous systems could also alter merger control.

Imagine several AI-managed businesses automatically acquire:

• startups;
• datasets;
• patents;
• cloud capacity;
• competitors;
• distribution networks.

If acquisition decisions are executed automatically, authorities must determine:

• who is acquiring whom;
• when control changes;
• whether the transaction constitutes a concentration;
• whether multiple transactions form a single economic strategy.

This creates the possibility of algorithmic serial acquisitions.

25. Autonomous Vertical Integration

A single autonomous system could progressively internalize functions.

For example:

marketplace → payment → logistics → cloud → AI → advertising

The system could therefore recreate the economic advantages of vertical integration without conventional corporate acquisitions.

Competition concerns may include:

• foreclosure;
• self-preferencing;
• tying;
• bundling;
• discriminatory access;
• raising rivals' costs;
•