Competition Law and Long-Term Evolution of Competition Policy in Post-Human Economies

1. Introduction

Competition policy in post-human economies refers to the future development of competition-law principles where economic activity is increasingly performed or controlled by artificial intelligence, autonomous agents, robotics, algorithmic systems, decentralized networks, machine-to-machine transactions, and other non-human decision-making systems.

The term “post-human economy” does not necessarily mean an economy without human beings. It describes an economic environment in which humans may no longer be the only—or even the primary—actors making commercial decisions.

Examples include:

AI purchasing agents;

autonomous trading systems;

robotic manufacturers;

AI-controlled supply chains;

autonomous logistics networks;

algorithmic marketplaces;

machine-to-machine contracting;

AI-generated products and services;

decentralized digital ecosystems;

autonomous financial systems.

The fundamental competition-law question becomes:

How should competition law evolve when important competitive decisions are increasingly made by autonomous systems rather than directly by human managers?

2. Meaning of a Post-Human Economy

A post-human economy may contain several layers:

Human layer

Consumers, entrepreneurs, investors and regulators.

Algorithmic layer

Pricing, advertising, recommendation and allocation algorithms.

AI layer

Systems capable of learning, predicting and making commercial decisions.

Autonomous-agent layer

AI agents capable of negotiating, purchasing, selling and contracting.

Robotic layer

Machines capable of production, logistics and physical distribution.

Infrastructure layer

Cloud computing, telecommunications, energy, semiconductor and data infrastructure.

These layers can interact continuously.

For example:

AI agent → searches marketplace → selects supplier → negotiates price → orders product → autonomous logistics → payment system → delivery robot.

Competition law must therefore regulate competition across the entire economic chain.

3. Why Competition Policy Must Evolve

Traditional competition law was developed largely around human organizations.

A conventional model is:

Company → manager → employee → decision → market conduct.

A future model may be:

AI → algorithm → autonomous agent → transaction → market outcome.

This creates new questions.

For example:

Who made the competitive decision?

Can an AI agent form an unlawful agreement?

Can autonomous algorithms coordinate prices?

Can an AI system discriminate against competitors?

Who is responsible for algorithmic exclusion?

Can an AI ecosystem become dominant?

How should merger control treat AI startups?

Can machines themselves be treated as economic actors?

4. Fundamental Objective

The purpose of competition law should remain the protection of the competitive process.

Core objectives include:

preventing unlawful collusion;

controlling abusive market power;

protecting contestability;

preserving innovation;

preventing unjustified foreclosure;

protecting consumer choice;

maintaining opportunities for entry.

The technology may change, but these fundamental objectives remain relevant.

5. From Human Competition to Machine-Mediated Competition

Traditional competition:

Human Firm A ↔ Human Firm B

Digital competition:

Platform A ↔ Platform B

AI competition:

AI System A ↔ AI System B

Autonomous competition:

AI Agent A ↔ AI Agent B

Post-human economic competition may eventually involve:

Autonomous Ecosystem A ↔ Autonomous Ecosystem B.

This progression creates a need to adapt competition-law analysis.

6. Market Definition in Post-Human Economies

Market definition may become more difficult.

Traditional questions include:

What products compete?

What geographic market exists?

Can consumers substitute one product for another?

Future markets may involve:

AI assistants;

autonomous agents;

cloud systems;

digital twins;

robotics;

decentralized networks.

A consumer may not directly choose a product.

Instead:

Consumer → AI Agent → Marketplace → Supplier

The AI agent may effectively determine which products receive demand.

Therefore, competition authorities may need to examine both:

consumer-facing markets; and

agent-mediated markets.

7. Dynamic Market Definition

Future markets can change extremely quickly.

An AI technology that appears insignificant today may become a major competitive constraint tomorrow.

Competition policy should therefore examine:

current competitors;

potential competitors;

emerging technologies;

innovation pipelines;

technological substitutes.

This is particularly important for merger control.

8. AI as a Source of Market Power

AI may generate market power through:

data;

computing capacity;

proprietary models;

infrastructure;

distribution;

talent;

network effects.

An AI ecosystem may therefore develop a feedback loop:

More users → more data → better AI → more users → more data.

Competition policy must determine whether such advantages arise through legitimate innovation or are reinforced by exclusionary conduct.

9. Autonomous Agents

An autonomous commercial agent could:

search for suppliers;

compare prices;

negotiate contracts;

purchase goods;

choose payment systems;

arrange logistics.

Suppose millions of consumers use AI purchasing agents.

The agent may become an important gatekeeper between consumers and suppliers.

Competition issues may arise if one agent:

favours affiliated suppliers;

excludes competing suppliers;

imposes discriminatory conditions;

charges excessive commissions;

prevents competing agents from accessing users.

10. Algorithmic Collusion

One major future concern is coordination through algorithms.

Traditional cartel:

Human executives communicate → agree on price → implement agreement.

Algorithmic coordination could potentially occur through:

Algorithms observe market behaviour → adapt prices → reach coordinated outcomes.

However, parallel algorithmic behaviour alone does not automatically establish an unlawful cartel.

Competition authorities still need to apply the relevant legal standards concerning:

agreement;

concerted practice;

communication;

conscious coordination;

facilitating practices;

market effects.

11. Autonomous Pricing

AI systems may adjust prices continuously.

For example:

Demand increases → AI raises price

and

Demand falls → AI reduces price.

This is normally legitimate dynamic pricing.

Competition concerns may arise if systems are designed or used to facilitate unlawful coordination or exclusion.

The legal challenge is distinguishing:

independent optimization

from

unlawful coordination.

12. Self-Learning Algorithms

A self-learning system can modify its behaviour without direct human instructions.

This creates an important legal question:

Can competition law attribute the system's conduct to the undertaking that designed, deployed or controlled it?

Existing competition law generally focuses on the conduct of undertakings rather than granting autonomous AI systems independent legal responsibility.

Thus, future policy may need clearer rules concerning:

design responsibility;

deployment responsibility;

monitoring duties;

auditability;

compliance controls.

13. The Problem of Human Intent

Traditional cartel cases often examine human evidence such as:

emails;

meetings;

telephone calls;

instructions.

AI systems may produce competitive outcomes without a conventional human agreement.

Future enforcement may therefore rely more heavily on:

system architecture;

training objectives;

code;

communication protocols;

data flows;

audit logs;

deployment instructions;

algorithmic interactions.

14. Network Effects

Network effects may become even stronger in post-human economies.

For example:

More AI agents → more transactions → more data → better AI → more agents.

This can cause markets to “tip” toward a small number of ecosystems.

Competition policy should therefore monitor:

user concentration;

agent concentration;

data advantages;

interoperability;

switching costs.

15. Data as Competitive Infrastructure

Data may become equivalent to critical infrastructure in some AI markets.

Competitive advantages may arise from:

exclusive datasets;

transaction histories;

consumer preferences;

industrial information;

real-time behavioural data.

Competition authorities may therefore examine whether control over data:

creates entry barriers;

prevents rivals from developing;

reinforces dominance;

facilitates discrimination.

16. Cloud and Compute Concentration

Advanced AI may depend upon:

cloud computing;

GPUs;

specialized chips;

data centres;

energy.

If these inputs are highly concentrated, competition at the AI application level may also become constrained.

A future competition framework may therefore examine:

Compute → AI models → applications → distribution → consumers.

17. Interoperability

Interoperability allows different AI systems to communicate.

For example:

AI Agent A ↔ AI Agent B

instead of:

AI Agent A → closed ecosystem.

Interoperability can reduce:

switching costs;

ecosystem lock-in;

network-effect barriers.

However, mandatory interoperability can also create:

security risks;

privacy risks;

innovation concerns.

Therefore, remedies must be proportionate.

18. Data Portability

Data portability may allow users to move:

Consumer data → AI Agent A → AI Agent B.

This can increase contestability.

But portability must be designed carefully because data may contain:

personal information;

confidential information;

trade secrets;

third-party rights.

Competition law must therefore interact with privacy and data-protection law.

19. Autonomous Ecosystems

Future businesses may become integrated ecosystems.

Example:

AI model + cloud + marketplace + payment + logistics + advertising + robotics.

This creates potential leverage.

A firm dominant in one layer could potentially use that dominance to strengthen another.

This resembles traditional vertical integration but can occur across much larger technological networks.

20. Self-Preferencing by AI

An AI platform may rank:

its own products;

affiliated services;

independent competitors.

If the ranking system systematically favours affiliated businesses, competition concerns may arise.

The Google Shopping proceedings provide an important reference point for understanding the relationship between dominance in a digital infrastructure layer and preferential treatment of affiliated services.

21. Tying and Bundling

Future AI ecosystems may combine:

AI assistant;

cloud;

search;

operating system;

payment;

marketplace.

A dominant undertaking may potentially condition access to one service on use of another.

Competition law must distinguish:

efficient integration

from

anticompetitive foreclosure.

22. Essential Facilities

Autonomous economies may depend upon critical infrastructure.

Examples include:

cloud infrastructure;

payment networks;

computing capacity;

communication networks;

AI interfaces.

The Bronner doctrine is important because it illustrates that competition law does not automatically require dominant firms to share every asset.

The strict legal conditions for compulsory access remain important.

23. Merger Control in Post-Human Economies

Merger control may become one of the most important areas of future competition policy.

Authorities may need to examine acquisitions involving:

AI startups;

robotics companies;

data companies;

autonomous-agent developers;

semiconductor businesses;

cloud infrastructure.

A target's current revenue may not adequately represent its future competitive significance.

24. Potential Competition

Suppose a dominant AI company acquires a startup with:

little revenue;

few users;

innovative technology;

strong research capability.

Traditional market-share analysis may underestimate the importance of the startup.

The authority may therefore examine whether the startup represented a potential competitive constraint.

25. Innovation Competition

Innovation may become more important than price competition.

Two AI companies may compete by:

improving accuracy;

reducing computing requirements;

improving safety;

developing new architectures;

creating better autonomous agents.

Consequently, competition policy should consider:

innovation competition, not merely price competition.

26. Robotics and Competition

Robotics may transform manufacturing and logistics.

Suppose one undertaking controls:

robot operating systems;

robotic hardware;

logistics software;

warehouses.

It could potentially influence competition throughout the supply chain.

Competition authorities may therefore need ecosystem-wide analysis.

27. Decentralized Economies

Blockchain and decentralized networks may change traditional assumptions about market structure.

There may be:

no conventional central company;

automated smart contracts;

decentralized governance;

autonomous protocols.

Competition law must determine:

who constitutes the undertaking;

who controls the protocol;

who benefits economically;

who can modify the system;

who can coordinate conduct.

28. Case Law

Case 1 — United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001)

Microsoft is a foundational technology-platform competition case.

Importance

The case demonstrates how a dominant position in one technological layer can be used to influence competition in adjacent markets.

Post-Human Relevance

The same reasoning may become relevant where an undertaking controls:

AI infrastructure → operating system → applications → distribution.

The lesson is that competition analysis must consider technological ecosystems rather than isolated products.

29. Case 2 — Ohio v. American Express Co., 585 U.S. 529 (2018)

American Express involved a two-sided transaction platform.

Importance

The case demonstrates that platform markets may involve interconnected groups of users.

Post-Human Relevance

Future autonomous markets may have several interacting sides:

Consumers ↔ AI agents ↔ Sellers ↔ Payment providers.

Competition effects may therefore need to be analyzed across interconnected market participants.

30. Case 3 — United Brands Co. v Commission, Case 27/76

United Brands is a classic European competition case concerning dominance.

Importance

It establishes foundational principles concerning:

relevant market;

dominance;

market power;

abusive conduct.

Post-Human Relevance

Even where markets are controlled by advanced AI systems, the fundamental question remains:

Does the undertaking possess substantial market power capable of affecting competitive conditions?

31. Case 4 — AKZO Chemie BV v Commission, Case C-62/86

AKZO is a major authority concerning predatory pricing.

Post-Human Relevance

Autonomous systems could potentially implement highly sophisticated pricing strategies.

Competition authorities must distinguish:

legitimate algorithmic pricing

from

pricing designed or used to exclude competitors unlawfully.

The case therefore remains relevant to future algorithmically controlled markets.

32. Case 5 — Bronner v Mediaprint, Case C-7/97

Bronner concerned access to a distribution system.

Importance

The case establishes important limitations concerning compulsory access to infrastructure.

Post-Human Relevance

Autonomous economies may depend upon:

cloud infrastructure;

AI infrastructure;

digital distribution systems.

Bronner illustrates that access obligations require satisfaction of the relevant legal conditions rather than merely demonstrating that access would benefit competition.

33. Case 6 — MOTOE v Elliniko Dimosio, Case C-49/07

MOTOE concerned an entity exercising regulatory powers while also participating in economic activity.

Importance

The case highlights the competitive risks arising where regulatory and commercial functions coexist.

Post-Human Relevance

Future autonomous ecosystems may have entities that:

establish technical standards;

control access;

operate infrastructure;

compete commercially.

The regulator + infrastructure operator + competitor combination can create important competition concerns.

34. Case 7 — Google Shopping Proceedings

The Google Shopping proceedings concern Google's treatment of comparison-shopping services within its search ecosystem.

Importance

The matter is particularly relevant to:

self-preferencing;

ranking;

platform dominance;

access to digital consumers.

Post-Human Relevance

Future AI assistants may decide which products consumers see.

Therefore:

AI recommendation + dominant platform + preferential ranking

could become a major competition-policy issue.

35. Case 8 — Google Android, Case AT.40099

The Google Android proceedings concerned contractual arrangements surrounding Android and related services.

Relevance

The matter illustrates how control of one ecosystem layer can influence competition in adjacent markets.

Post-Human Relevance

The same issue may arise in:

AI operating systems;

AI assistants;

autonomous-agent marketplaces;

robotic operating systems.

36. Case 9 — Intel Corp. v European Commission, Case C-413/14 P

Intel is important in the law concerning exclusionary rebates and assessment of competitive effects.

Post-Human Relevance

Future autonomous platforms may use sophisticated incentive systems to encourage users or businesses to remain within their ecosystem.

Competition authorities may therefore need to examine whether:

rebates;

discounts;

incentives;

loyalty mechanisms

foreclose equally efficient competitors.

37. Case 10 — Eturas UAB v Lietuvos Respublikos konkurencijos taryba, Case C-74/14

Eturas concerned an online booking platform and the transmission of information capable of facilitating coordinated pricing behaviour.

Importance

It is especially relevant to digital-platform competition.

Post-Human Relevance

It illustrates how technology-mediated communications can become relevant to competition-law analysis.

Future systems may replace human communications with:

APIs;

automated signals;

machine-readable instructions;

AI-to-AI communications.

38. Algorithmic Collusion: Future Legal Framework

Future competition policy may need to distinguish four situations.

Situation 1 — Independent algorithms

Each firm independently chooses its strategy.

Generally, algorithmic independence does not itself establish a cartel.

Situation 2 — Human-directed coordination

Managers instruct algorithms to implement an agreement.

Traditional cartel principles may apply.

Situation 3 — Algorithm-assisted coordination

Algorithms facilitate communication or implementation of an unlawful arrangement.

Traditional competition principles may still apply depending on the facts.

Situation 4 — Autonomous coordination

Independent systems unexpectedly converge on similar strategies.

This presents a more difficult legal problem because coordination may occur without conventional human communication.

39. Liability and Accountability

A major future issue is attribution.

Possible responsible parties include:

AI developer;

platform operator;

deploying company;

algorithm owner;

controlling undertaking;

participating businesses.

A possible regulatory principle is:

Autonomous technology should not become a mechanism for avoiding responsibility for conduct attributable to an undertaking under competition law.

40. Competition Policy and Human Oversight

Even highly autonomous markets may require human oversight.

Companies may need:

algorithmic compliance systems;

competition-risk assessments;

audit trails;

monitoring;

internal controls;

escalation mechanisms.

This can make competition compliance part of AI governance.

41. Long-Term Evolution of Competition Policy

The evolution may occur in several stages.

Stage 1 — Traditional competition law

Focus:

Human agreements + human firms.

Stage 2 — Digital competition law

Focus:

Platforms + data + network effects.

Stage 3 — Algorithmic competition law

Focus:

Algorithms + automated pricing + digital coordination.

Stage 4 — Autonomous competition law

Focus:

AI agents + autonomous transactions + machine ecosystems.

Stage 5 — Post-human competition policy

Focus:

Interconnected autonomous economic systems and long-term ecosystem governance.

42. Ex-Ante and Ex-Post Regulation

Future competition policy may require both.

Ex-post

Used after suspected anti-competitive conduct occurs.

Examples:

cartel enforcement;

abuse-of-dominance cases;

exclusionary conduct;

merger review.

Ex-ante

Rules may apply before harm becomes entrenched.

Potential areas:

interoperability;

portability;

gatekeeper obligations;

access;

transparency;

ecosystem conduct.

The objective is to prevent markets from becoming practically impossible to reopen.

43. Institutional Challenges

Competition authorities may need new technical capabilities.

They may require:

AI specialists;

software engineers;

economists;

data scientists;

cybersecurity experts;

algorithm auditors.

Traditional legal analysis alone may not reveal how an autonomous system operates.

44. Evidence in Autonomous Markets

Evidence may include:

source code;

model documentation;

training data;

system logs;

API records;

algorithmic instructions;

transaction records;

communications between agents.

This creates interaction between competition law and digital evidence law.

45. Consumer Welfare

Consumer welfare remains relevant.

Authorities may examine:

price;

quality;

choice;

innovation;

privacy;

security;

convenience.

In AI markets, a service may be free but still generate competition concerns through:

data extraction;

reduced choice;

exclusion;

declining quality.

46. Competition and Privacy

Competition and privacy can overlap.

For example:

Dominant platform → exclusive control of data → superior AI → stronger dominance.

However, privacy protection and competition law have different legal objectives.

They should therefore be coordinated without treating them as identical fields.

47. Competition Neutrality

Post-human economies may contain government-supported AI infrastructure.

Competition policy should consider whether public support gives particular firms unjustified advantages.

This is particularly relevant to:

cloud infrastructure;

AI compute;

semiconductor manufacturing;

strategic technology.

48. International Cooperation

Autonomous systems operate across borders.

A single AI platform could simultaneously affect:

India;

Europe;

the United States;

the Middle East;

Southeast Asia.

Competition authorities may therefore need:

international information sharing;

coordinated merger review;

compatible remedies;

technical cooperation.

49. Future Remedies

Potential remedies may include:

Structural remedies

Separation or divestiture where legally justified.

Behavioural remedies

Restrictions on:

exclusivity;

self-preferencing;

tying;

discriminatory access.

Technical remedies

interoperability;

API access;

portability.

Merger remedies

Conditions preventing elimination of emerging competitors.

Monitoring remedies

Continuous supervision of high-risk ecosystems.

50. Major Challenges

Competition law in post-human economies will face several challenges.

1. Attribution

Who is responsible for autonomous conduct?

2. Explainability

How can authorities understand an AI decision?

3. Speed

Algorithms can change behaviour faster than investigations.

4. Scale

One system can affect millions of transactions simultaneously.

5. Cross-border operation

Autonomous systems can operate globally.

6. Innovation

Excessive intervention may discourage beneficial technological development.

7. Uncertainty

Future competitive effects may be difficult to predict.

51. Practical Example

Imagine AutoMarket AI, an autonomous commercial ecosystem.

It:

selects suppliers;

negotiates prices;

manages warehouses;

controls logistics;

operates payment systems;

recommends products;

manages advertising;

uses proprietary AI.

It then acquires three promising AI startups.

Its algorithm gives its own products preferential ranking.

Its contracts prevent sellers from using competing AI agents.

A future competition authority might ask:

Does AutoMarket possess dominance?

What is the relevant market?

Does its AI create network effects?

Does it control essential infrastructure?

Does self-preferencing disadvantage competitors?

Do exclusivity provisions prevent multi-homing?

Do acquisitions eliminate potential competition?

Does control over data create durable entry barriers?

Can sellers switch to rival AI systems?

Can effective competition re-emerge if competitors are excluded?

52. Principles for Future Competition Policy

A long-term framework should incorporate:

Principle 1 — Technology neutrality

Competition rules should apply regardless of whether conduct is performed by humans or algorithms.

Principle 2 — Accountability

Autonomous technology should not eliminate legal responsibility.

Principle 3 — Contestability

Markets should remain open to new competitors.

Principle 4 — Innovation

Competition policy should protect future innovation.

Principle 5 — Interoperability

Appropriate interoperability can prevent excessive lock-in.

Principle 6 — Data mobility

Appropriate portability can reduce switching costs.

Principle 7 — Dynamic merger review

Potential competition should be considered where legally relevant.

Principle 8 — Proportionality

Intervention should correspond to demonstrated competition concerns.

53. Long-Term Policy Formula

Post-Human Competition Policy =

Market Power + AI Accountability + Algorithmic Coordination Control + Data Governance + Interoperability + Portability + Innovation Protection + Dynamic Merger Control + Ecosystem Monitoring + Proportionate Remedies

54. Conclusion

The long-term evolution of competition policy in post-human economies will involve adapting established competition principles to markets in which AI, algorithms, autonomous agents, robotics and interconnected technological ecosystems increasingly make economic decisions.

The central legal principles remain recognizable:

dominance must be assessed;

anti-competitive agreements must be prevented;

exclusionary conduct must be examined;

market access must remain meaningful;

innovation must be protected;

future competitors must not unnecessarily be eliminated.

Cases such as Microsoft, United Brands, AKZO, Bronner, American Express, MOTOE, Google Shopping, Google Android, Intel and Eturas provide different building blocks for this evolution.

The fundamental challenge is not simply whether machines can compete with humans. It is whether competition law can preserve independent rivalry, innovation and contestability when economic ecosystems themselves become increasingly autonomous.

Quick Revision Formula

Post-Human Competition Policy = Autonomous Markets + AI Accountability + Algorithmic Coordination + Network Effects + Data + Interoperability + Innovation + Potential Competition + Merger Control + Ecosystem Governance + Proportionate Remedies.

One-Line Definition

Competition policy in post-human economies is the evolving application of competition-law principles to economic systems in which AI, autonomous agents, algorithms, robotics and interconnected technological ecosystems increasingly perform competitive and commercial functions.