Competition Law and Strategic Foresight for Competition Governance in Computational Civilizations

1. Introduction

Strategic foresight for competition governance in computational civilizations refers to the use of competition law, economic analysis, technological monitoring and forward-looking regulatory tools to anticipate how AI, algorithms, autonomous systems, cloud infrastructure, data networks, digital platforms and computational ecosystems may reshape market power before conventional antitrust enforcement becomes too slow.

The expression “computational civilization” may be understood as an economy in which increasingly important commercial decisions are made or assisted by:

• artificial intelligence;
• algorithmic pricing;
• autonomous agents;
• recommendation and ranking systems;
• large-scale data processing;
• cloud and computing infrastructure;
• digital identity and payment systems;
• foundation models;
• automated procurement;
• platform ecosystems; and
• machine-to-machine transactions.

The OECD has specifically identified algorithmic collusion, algorithmic exclusion and exploitative algorithmic conduct as emerging competition concerns, while also recognising that algorithms can produce substantial efficiency and innovation benefits.

Thus, competition governance must move from merely asking “Has an antitrust violation already occurred?” to also asking:

“What computational architecture is likely to create durable market power, coordination or dependency in the future?”

This is the central function of strategic foresight.

2. Meaning of Strategic Foresight in Competition Law

Strategic foresight is not ordinary prediction. It involves systematically examining:

1. technological trends;
2. emerging market structures;
3. possible future theories of harm;
4. alternative regulatory scenarios;
5. early-warning indicators;
6. potential concentration of infrastructure;
7. switching and interoperability barriers; and
8. possible effects of intervention or non-intervention.

In computational markets, foresight is particularly important because market power can develop extremely rapidly.

A company may initially appear to compete in one market but gradually become indispensable across several layers:

Compute → Cloud → Data → Foundation Model → Application → Distribution → Consumer Interface

Control of several layers can create ecosystem power.

3. Competition Governance in a Computational Civilization

Traditional competition law generally operates through three principal mechanisms:

A. Ex-post enforcement

Authorities investigate conduct after it occurs.

Examples:

• cartel enforcement;
• abuse of dominance;
• exclusionary agreements;
• predatory conduct;
• discriminatory access;
• anticompetitive mergers.

B. Merger control

Authorities examine whether a transaction could substantially reduce future competition.

This becomes particularly important for:

• AI startups;
• cloud providers;
• semiconductor firms;
• data platforms;
• foundation-model developers;
• cybersecurity providers;
• digital infrastructure.

C. Ex-ante regulation

Newer digital-market regimes can impose obligations before conventional antitrust harm becomes fully established.

The EU Digital Markets Act is an important example. The Commission has used it to examine matters such as app-store steering, self-preferencing and consumer choice.

Strategic foresight connects all three mechanisms.

4. Why Computational Markets Require Foresight

4.1 Speed of market evolution

Digital markets can evolve faster than litigation.

A conventional antitrust investigation may take years, while an AI ecosystem can change significantly during that period.

Consequently:

Slow enforcement + rapid technological change = risk of obsolete remedies.

4.2 Network effects

The value of a computational platform may increase as more participants use it.

Examples include:

• social platforms;
• payment networks;
• app stores;
• cloud ecosystems;
• AI developer platforms;
• digital marketplaces.

Network effects can produce self-reinforcing market structures.

4.3 Data advantages

Large datasets can improve:

• prediction;
• recommendation;
• advertising;
• fraud detection;
• AI training;
• personalization.

A dominant firm may therefore obtain a data feedback loop:

More users → more data → better algorithms → better service → more users → more data.

Competition governance must determine when such feedback represents legitimate innovation and when it becomes an exclusionary mechanism.

5. Algorithmic Coordination

One of the most important foresight problems is algorithmic collusion.

Algorithms can:

• observe competitors' prices;
• respond instantaneously;
• adjust prices repeatedly;
• predict competitor behaviour;
• exchange information indirectly;
• optimise for long-term profit.

The OECD has warned that algorithms can potentially facilitate coordination even where traditional forms of communication are absent, although the legal treatment depends on the circumstances and applicable jurisdiction.

The critical distinction is between:

Legitimate independent algorithmic pricing

Each undertaking independently develops an algorithm to maximise its own commercial objectives.

and

Anticompetitive algorithmic coordination

Algorithms are deliberately designed, supplied or used to facilitate:

• price fixing;
• market allocation;
• information exchange;
• output restriction;
• coordinated exclusion.

The computational mechanism does not automatically determine legal liability. The underlying conduct and evidence remain crucial.

6. Algorithmic Self-Preferencing

A computational platform may simultaneously act as:

1. infrastructure provider;
2. marketplace operator;
3. search engine;
4. data intermediary; and
5. competitor to businesses using its infrastructure.

This creates a structural conflict.

For example:

Platform → collects market data → identifies successful rival → changes ranking algorithm → promotes its own competing service.

This is the conceptual problem of algorithmic self-preferencing.

The Google Shopping litigation is particularly important here.

7. Case Laws

Case 1: Google and Alphabet v European Commission — Google Shopping

Case C-48/22 P, judgment of 10 September 2024

The Court of Justice upheld the finding concerning Google's preferential positioning of its own comparison-shopping service in general search results. The Court confirmed the Commission's finding of abuse of dominance and upheld the approximately €2.4 billion fine.

Importance

This case is highly relevant to computational competition governance because ranking algorithms can function as a competitive gatekeeping mechanism.

The central lesson is:

Competition law can examine how a dominant computational intermediary structures visibility and access to consumers.

Foresight significance

Future investigations may need to examine:

• AI-generated rankings;
• recommendation engines;
• search-result ordering;
• marketplace visibility;
• automated eligibility systems;
• algorithmic allocation of consumer attention.

The important issue is therefore not merely who owns the algorithm, but whether algorithmic architecture is being used to distort competitive opportunities.

Case 2: Eturas and Others v Lietuvos Respublikos konkurencijos taryba

Case C-74/14, judgment of 21 January 2016

Eturas concerned travel agencies using a common computerised booking system. The system administrator communicated a restriction affecting online discounts, and the case considered whether the circumstances could establish a concerted practice.

Importance

Eturas demonstrates that digital systems can become the mechanism through which coordination occurs.

The important legal question is not whether the communication was physically delivered through a computer system, but whether the evidence establishes the necessary elements of coordination.

Foresight significance

This principle becomes increasingly important where:

• competitors use a common software provider;
• software automatically implements commercial restrictions;
• competitors receive common algorithmic signals;
• platform architecture standardises competitive behaviour.

It therefore provides an early legal foundation for analysing machine-mediated coordination.

Case 3: Intel Corp. v Commission

Case C-413/14 P; T-286/09 RENV

Intel concerned conditional rebates and alleged exclusionary conduct in the x86 microprocessor market.

In 2022, following the earlier appellate proceedings and remittal, the General Court annulled part of the Commission's decision because the Commission had not sufficiently established the anticompetitive capability of the rebates under the applicable effects analysis.

Importance for computational civilization

Intel demonstrates an essential principle of technologically sophisticated competition enforcement:

Market power alone is not enough; the authority must properly establish the competitive effects of the challenged conduct where the applicable legal framework requires such analysis.

Foresight significance

This matters for AI and computational ecosystems because regulators may encounter:

• preferential cloud pricing;
• AI-compute rebates;
• exclusive developer incentives;
• bundled access to models;
• loyalty discounts for infrastructure;
• conditional access to APIs.

Strategic foresight must therefore anticipate potential foreclosure without abandoning rigorous economic and legal analysis.

Case 4: Google Android — AT.40099

The European Commission's 2018 Android decision concerned Google's conduct relating to Android devices, including restrictions connected with search and app ecosystems. The Commission imposed a €4.34 billion fine.

Importance

Android illustrates ecosystem leverage.

A company may use control over one technological layer to influence competition at another layer.

Conceptually:

Operating system → app distribution → search → advertising → data

The competitive concern is therefore not limited to an individual product.

Foresight significance

Future computational ecosystems may involve:

Cloud → AI model → operating environment → applications → data

Competition authorities should therefore examine cross-layer leverage rather than analysing every product as an isolated market.

Case 5: United States v. Microsoft Corp.

The Microsoft litigation remains a foundational case for understanding technological ecosystem power.

Microsoft's conduct concerning the Windows operating-system ecosystem and browser competition demonstrated how control over an important technological platform can influence adjacent markets.

Importance

The case established an enduring competition-law lesson:

A dominant technological platform may possess the ability to influence the competitive conditions under which complementary technologies develop.

Computational relevance

The principle can be applied conceptually to:

• AI operating environments;
• cloud platforms;
• foundation models;
• application marketplaces;
• digital assistants;
• developer APIs.

Strategic foresight therefore asks whether today's platform advantage can become tomorrow's ecosystem bottleneck.

Case 6: United States v. Apple Inc.

The U.S. Department of Justice and state plaintiffs filed the smartphone monopolization case against Apple in March 2024. The complaint alleges that Apple used contractual restrictions and control over technical access points, including APIs and app distribution, to maintain its position and limit competitive threats. The litigation remains a live matter rather than a final merits judgment.

Importance

This case is especially relevant to computational competition governance because it illustrates the relationship between:

• interoperability;
• APIs;
• switching costs;
• ecosystem control;
• technical restrictions;
• complementary innovation.

Foresight significance

Tomorrow's equivalent question could concern:

• AI-agent interoperability;
• cross-model portability;
• access to foundation-model APIs;
• cloud switching;
• autonomous-agent identity;
• data portability.

Competition governance must therefore treat interoperability as a potential competitive variable.

Case 7: Online Poster Price-Fixing — Trod Ltd.

The U.S. Department of Justice prosecuted Trod and an executive in connection with price fixing involving posters sold through Amazon Marketplace. The conduct involved online pricing and illustrates the use of automated or algorithmic mechanisms to implement coordinated pricing.

Importance

This is a practical example of an important principle:

Technology does not make an otherwise unlawful cartel lawful.

If competitors agree to coordinate and then use software to execute that agreement, the algorithm is merely the technological implementation mechanism.

Foresight significance

Future enforcement may need to investigate:

• common repricing software;
• AI pricing agents;
• common optimisation providers;
• shared market intelligence;
• automated responses to competitors.

The OECD's more recent work specifically identifies shared pricing software and common model providers as potential areas of concern.

8. Strategic Foresight: Emerging Competition Risks

A. AI Foundation Models

A future competition authority may need to examine whether control over:

• compute;
• training data;
• specialised chips;
• model weights;
• distribution;
• cloud infrastructure

creates cumulative market power.

The important question is not simply:

“Who has the largest AI model?”

It is:

“Which combination of computational assets makes meaningful entry increasingly difficult?”

B. Autonomous AI Agents

AI agents may eventually negotiate:

• prices;
• procurement;
• contracts;
• transportation;
• insurance;
• financial services.

This creates a new competition question:

If machines negotiate with other machines, who is legally responsible for the resulting coordination?

Possible actors include:

• the user;
• the firm deploying the agent;
• the developer;
• the model provider;
• the platform;
• the intermediary supplying market information.

This will create difficult attribution questions.

9. Algorithmic Governance and Evidence

Competition authorities increasingly need technical evidence such as:

• source-code documentation;
• model cards;
• audit logs;
• training-data information;
• API records;
• version histories;
• algorithmic decision logs;
• A/B testing records;
• pricing histories;
• ranking changes;
• internal communications.

The OECD has emphasised that sophisticated technical tools can be useful, but authorities do not necessarily need highly complex technical techniques in every case; appropriate evidence is case-specific.

Thus:

Computational competition law ≠ purely technical competition law.

Traditional evidence remains important.

10. Predictive Competition Monitoring

Strategic foresight could create an early-warning system based on indicators such as:

These are screening indicators, not proof of infringement.

11. Competition Governance Architecture

A future-oriented framework can be represented as follows:

TECHNOLOGICAL FORESIGHT          ↓ Identify emerging computational systems          ↓ Map ecosystem structure          ↓ Identify control points / bottlenecks          ↓ Measure market power          ↓ Identify potential theories of harm          ↓ Continuous data and algorithmic monitoring          ↓ ┌──────────────────────────────┐ │ Competition intervention     │ ├──────────────────────────────┤ │ Ex-post enforcement          │ │ Merger control               │ │ Ex-ante regulation           │ │ Interoperability remedies    │ │ Access remedies              │ │ Behavioural commitments      │ │ Structural remedies          │ └──────────────────────────────┘          ↓ Monitor market evolution          ↓ Update regulatory strategy

12. From Market Definition to Ecosystem Mapping

Traditional competition analysis frequently begins with:

Relevant product market + relevant geographic market.

Computational ecosystems may require a broader analytical map.

For example:

Semiconductors      ↓ Computing infrastructure      ↓ Cloud      ↓ Foundation models      ↓ AI applications      ↓ Distribution platforms      ↓ Consumers

A competition authority should examine whether control at one layer allows a firm to influence competitive conditions at another.

This is particularly important because computational markets frequently involve multi-sided platforms and complementary products.

13. Strategic Foresight and Merger Control

Traditional merger control asks whether a proposed concentration is likely to substantially lessen competition or otherwise satisfy the applicable jurisdictional test.

Computational markets create additional questions:

1. Does the target possess future strategic importance?

A small AI company may have:

• valuable researchers;
• unique datasets;
• specialised models;
• critical algorithms;
• important developer communities.

2. Is the target an emerging competitor?

3. Could acquisition eliminate future innovation?

4. Does the transaction combine complementary bottlenecks?

5. Could the transaction create vertical foreclosure?

6. Does it increase control over computational infrastructure?

Strategic foresight therefore extends merger analysis beyond current revenue and market shares.

14. Interoperability as a Competition Tool

Interoperability can reduce:

• switching costs;
• network-effect barriers;
• ecosystem lock-in;
• dependency on a single platform.

Possible future remedies include:

• API access;
• data portability;
• interoperability standards;
• technical compatibility;
• cross-platform messaging;
• model portability;
• cloud switching mechanisms.

The Apple litigation illustrates how technical access points and interoperability can become central to monopolization analysis.

15. Computational Essential Facilities

A future competition dispute may concern access to:

• AI compute;
• specialised chips;
• cloud infrastructure;
• critical datasets;
• interoperability interfaces;
• digital identity infrastructure;
• payment rails;
• model APIs.

The essential-facilities question must remain legally disciplined. Not every commercially useful computational resource becomes an essential facility.

The relevant inquiry may include:

1. indispensability;
2. lack of realistic alternatives;
3. competitive foreclosure;
4. justification for refusal;
5. proportionality of any access remedy.

16. Dynamic Market Power

Computational market power is frequently dynamic rather than static.

A firm can obtain advantages through:

Users → Data → Better model → Better product → More users

This creates a feedback loop.

Therefore, competition governance should examine:

Static power

Current market position.

Dynamic power

Ability to improve that position over time.

Architectural power

Ability to control the infrastructure through which competitors must operate.

Ecosystem power

Ability to influence several connected markets simultaneously.

17. Strategic Foresight and Competition Remedies

Traditional remedies may become inadequate if the computational environment changes rapidly.

Possible remedies include:

Behavioural remedies

• non-discrimination;
• transparency;
• restrictions on self-preferencing;
• fair access;
• algorithmic compliance.

Technical remedies

• interoperability;
• API access;
• data portability;
• technical separation;
• auditability.

Structural remedies

• divestiture;
• separation of business units;
• restrictions on vertical integration.

Procedural remedies

• continuous monitoring;
• periodic compliance reporting;
• independent technical audits.

The EU's experience with ex-ante digital regulation demonstrates the movement toward combining traditional competition enforcement with forward-looking obligations.

18. Competition Governance of AI Pricing

AI pricing introduces several possible scenarios:

Scenario 1 — Independent optimisation

Each firm independently optimises prices.

Potential effect: efficiency.

Scenario 2 — Algorithmic monitoring

Algorithms observe competitors and rapidly respond.

Potential issue: increased transparency and reduced competitive uncertainty.

Scenario 3 — Common pricing provider

Several competitors use the same external pricing system.

Potential issue: information exchange or coordinated pricing.

Scenario 4 — Explicit coordination through algorithms

Competitors intentionally use software to implement an agreement.

Potential issue: conventional cartel liability may apply despite technological implementation.

The OECD's work specifically distinguishes efficiency-enhancing algorithmic use from algorithmic mechanisms that restrict competition.

19. Computational Civilizations and Consumer Welfare

Competition governance should also examine non-price dimensions of competition.

Relevant variables include:

• privacy;
• quality;
• innovation;
• security;
• interoperability;
• choice;
• transparency;
• reliability.

In digital ecosystems, a consumer may pay zero monetary price while still facing competitive harm through:

• reduced quality;
• restricted choice;
• higher switching costs;
• reduced innovation;
• data exploitation.

Consequently, a purely price-centred approach may be insufficient.

20. International Cooperation

Computational ecosystems are inherently transnational.

An AI platform may:

• be incorporated in one jurisdiction;
• train models using data from another;
• use cloud infrastructure elsewhere;
• sell services globally;
• operate through subsidiaries across many jurisdictions.

Competition authorities therefore increasingly need cooperation concerning:

• evidence;
• merger review;
• algorithmic investigations;
• remedies;
• market studies;
• technical expertise.

The OECD has identified international coordination as an important component of competition policy for digital markets.

21. Key Legal Principles Emerging

From the cases and regulatory developments, several principles emerge.

Principle 1 — Algorithms do not create a separate antitrust universe

Existing competition principles can often apply to technologically implemented conduct.

Principle 2 — Technology can amplify market power

Control over computational infrastructure may strengthen existing dominance.

Principle 3 — Algorithmic conduct requires technical evidence

Traditional documents may need to be supplemented by computational evidence.

Principle 4 — Interoperability can be competitively significant

Technical restrictions may increase switching costs and protect ecosystem power.

Principle 5 — Self-preferencing can operate through ranking architecture

Google Shopping illustrates the relevance of algorithmic visibility and preferential positioning.

Principle 6 — Coordination can be machine-mediated

Eturas and online pricing cases demonstrate why competition authorities must examine digital mechanisms through which coordination occurs.

Principle 7 — Foresight cannot replace proof

Predicted future harm should not automatically be treated as established infringement.

22. Challenges for Competition Authorities

A. Black-box systems

Authorities may have difficulty understanding proprietary models.

B. Rapid technological change

The market may change before litigation concludes.

C. Attribution

It may be unclear whether conduct originated from:

• humans;
• AI developers;
• model providers;
• platforms;
• autonomous agents.

D. False positives

Not every concentration or algorithmic practice is anticompetitive.

E. Innovation risk

Over-regulation may discourage beneficial technological development.

F. Cross-border enforcement

Different jurisdictions may apply different standards.

23. A Strategic Foresight Model

A useful competition-governance model can be expressed as:

SCAN → MAP → MODEL → TEST → INTERVENE → MONITOR

SCAN

Identify emerging technologies.

MAP

Map firms, infrastructure, data, APIs and ecosystems.

MODEL

Develop possible future competitive scenarios.

TEST

Evaluate theories of harm using economic and legal evidence.

INTERVENE

Select appropriate enforcement or regulatory mechanisms.

MONITOR

Continuously reassess the market.

This creates a continuous competition-governance cycle rather than a one-time enforcement event.

24. Conclusion

Strategic foresight for competition governance in computational civilizations represents the evolution of competition law from a primarily reactive discipline toward a combination of enforcement, technological surveillance, economic analysis and institutional anticipation.

The major transformation is that competitive power may no longer reside solely in ownership of a product. It may reside in control over:

• computational infrastructure;
• data;
• algorithms;
• AI models;
• APIs;
• cloud systems;
• digital identity;
• distribution channels;
• interoperability standards;
• ecosystem interfaces.

The Google Shopping, Eturas, Intel, Google Android, Microsoft and Apple matters demonstrate different dimensions of this transformation: algorithmic ranking, computer-mediated coordination, exclusionary incentives, ecosystem leverage, platform power and interoperability.

The central proposition is therefore:

Competition law in a computational civilization must protect not merely today's competitive process, but the conditions that allow tomorrow's competitors, technologies and business models to emerge.