Competition Law and Adaptive Antitrust Systems for Autonomous Ecosystems

1. Introduction

Adaptive antitrust systems for autonomous ecosystems refers to competition-law frameworks capable of responding continuously to markets in which software, artificial intelligence, algorithms, platforms, connected devices and automated decision-making systems increasingly determine competitive conditions.

Traditional competition law generally examines a completed transaction or identifiable conduct: an agreement, exclusionary practice, tying arrangement, refusal to deal, merger or abuse of dominance. Autonomous ecosystems create a different problem. Competitive conditions may change automatically and continuously because algorithms can modify prices, rankings, access conditions, recommendations, interoperability, resource allocation and even contractual terms.

An autonomous ecosystem may therefore involve:

• an operating system;
• an app store;
• cloud infrastructure;
• AI models and agents;
• data repositories;
• APIs;
• payment systems;
• search and recommendation engines;
• automated pricing systems;
• connected devices;
• digital identity systems; and
• third-party developers and complementary businesses.

Competition law must consequently move from a purely event-based enforcement model toward a more continuous, data-driven and adaptive model.

Recent regulatory developments illustrate this direction. For example, the EU Digital Markets Act now imposes ex ante obligations on designated gatekeepers, while the European Commission has used binding specification measures concerning AI interoperability and access to Google Search data.

2. Meaning of an Autonomous Ecosystem

An autonomous ecosystem is a market environment in which significant competitive decisions are made or implemented through automated systems rather than exclusively through direct human decisions.

For example:

AI platform → operating system → app marketplace → payment system → ranking algorithm → recommendation engine → consumer → behavioural data → AI platform.

The ecosystem can continuously learn from the behaviour of its participants.

This creates a feedback loop:

Data → Algorithm → Decision → Market behaviour → New data → Algorithmic adaptation

Competition concerns may therefore arise even where no single human decision-maker explicitly intends to exclude competitors.

3. Why Traditional Antitrust Is Challenged

A. Dynamic market power

Traditional dominance analysis often asks whether an undertaking possesses substantial market power.

Autonomous ecosystems require additional questions:

1. Can the undertaking change competitive conditions automatically?
2. Can algorithms detect and respond to competitors in real time?
3. Can the platform alter ranking or access rules without renegotiating contracts?
4. Can AI systems learn strategies that disadvantage rivals?
5. Can the ecosystem use data obtained from dependent businesses to compete against them?

Market power may therefore become dynamic rather than static.

B. Algorithmic coordination

Algorithms can facilitate coordination between competitors.

The legal problem becomes:

When does autonomous algorithmic behaviour constitute independent parallel conduct, and when does it implement or facilitate an unlawful agreement?

The RealPage litigation illustrates this issue. The U.S. Department of Justice alleged that competing landlords supplied competitively sensitive information to RealPage's pricing system and used algorithmically generated recommendations, bringing Sections 1 and 2 of the Sherman Act into play.

The important principle is that technology does not automatically remove traditional antitrust liability.

4. Adaptive Antitrust: Core Principles

An adaptive competition regime should combine several mechanisms.

1. Continuous market monitoring

Authorities should monitor:

• prices;
• rankings;
• access conditions;
• API restrictions;
• switching costs;
• interoperability;
• data access;
• algorithmic changes;
• acquisitions;
• exclusionary patterns.

2. Algorithmic auditing

Competition authorities may need technical capacity to examine:

• training data;
• input variables;
• recommendation systems;
• pricing models;
• ranking systems;
• automated decision rules;
• model outputs;
• system logs.

3. Behavioural remedies

Instead of relying exclusively on fines, authorities may require:

• interoperability;
• data portability;
• API access;
• non-discriminatory ranking;
• interoperability testing;
• restrictions on self-preferencing;
• transparency obligations.

4. Structural remedies

Where behavioural remedies repeatedly fail, competition law may consider:

• separation of business units;
• divestiture;
• prohibition of particular acquisitions;
• restrictions on ecosystem integration.

5. Ex ante regulation

The EU DMA represents an important example of moving beyond conventional ex post Article 102 TFEU enforcement.

The DMA requires designated gatekeepers to comply with obligations concerning areas such as app distribution, steering and access conditions.

5. Six Major Case Laws

Case 1: Google LLC and Alphabet Inc. v European Commission — Google Android

Case: Google and Alphabet v European Commission, Case T-604/18, General Court, 14 September 2022.

The case concerned Google's Android ecosystem, including:

• Google Search;
• Chrome;
• Play Store;
• Android;
• device manufacturers;
• mobile network operators.

The General Court examined Google's contractual arrangements, including product bundling, exclusivity payments and anti-fragmentation obligations. The judgment expressly considered the concepts of multi-sided platforms and ecosystems.

Significance

The case demonstrates that competition law may examine an ecosystem as an interconnected system rather than analysing every product in complete isolation.

For autonomous ecosystems, the lesson is particularly important:

dominance in one technological layer can potentially reinforce power at other ecosystem layers.

Case 2: Google Search / Google Shopping

The Google Shopping litigation concerned Google's treatment of its own comparison-shopping service in search results.

The fundamental competition concern was self-preferencing: an ecosystem operator controlling an important gateway may have the ability to favour its own downstream service.

The concept becomes particularly significant in autonomous ecosystems because AI-driven ranking systems can automatically determine which competing products, services or applications receive visibility.

Adaptive-antitrust significance

Authorities may need to examine:

• ranking criteria;
• changes to algorithms;
• treatment of competing services;
• access to relevant data;
• interoperability;
• visibility allocation.

The European Commission has subsequently used the DMA to impose explicit obligations concerning self-preferencing. In July 2026, it announced a €460 million fine against Google concerning self-preferencing on Google Search.

Case 3: Qualcomm v European Commission

Case: Qualcomm v European Commission, Case T-235/18, General Court, 15 June 2022.

The case involved Qualcomm's exclusivity payments in the LTE chipset market and the assessment of exclusionary effects under Article 102 TFEU.

Relevance to autonomous ecosystems

Autonomous ecosystems frequently contain strategic bottlenecks such as:

• chips;
• operating systems;
• cloud infrastructure;
• AI accelerators;
• communications standards.

Control over a technologically indispensable layer can allow an undertaking to influence competition at downstream levels.

The case therefore illustrates the importance of examining foreclosure effects, rather than simply asking whether competitors technically remain able to enter the market.

Case 4: Qualcomm v European Commission — UMTS Chipsets

Case: Qualcomm v European Commission, Case T-671/19, General Court, 18 September 2024.

The case concerned alleged predatory pricing in the UMTS baseband chipset market. The General Court addressed relevant-market definition, dominance, price-cost analysis, exclusionary conduct and objective justification.

Relevance

Autonomous ecosystems can dynamically manipulate:

• prices;
• subsidies;
• commissions;
• access charges;
• cross-subsidies between ecosystem layers.

Consequently, antitrust authorities may need to examine algorithmically determined prices and cost structures, rather than merely posted prices.

Case 5: United States v RealPage

Case: United States and State Plaintiffs v RealPage, Inc.

This is particularly important for AI-driven competition.

The DOJ alleged that RealPage's pricing software enabled competing landlords to exchange and use competitively sensitive information through a centralized algorithmic system. The complaint invoked Sections 1 and 2 of the Sherman Act.

The litigation has continued through 2026, including settlements and proposed judgments involving participating property-management companies.

Principle

The key issue is not simply whether a computer generated the price.

The relevant questions include:

• Who supplied the data?
• Was the data competitively sensitive?
• What did participants agree to do with the system?
• Did participants knowingly rely on competitors' information?
• Did the algorithm facilitate coordinated outcomes?

Thus:

Algorithmic autonomy does not necessarily equal legal autonomy.

Human agreements surrounding the system may still create antitrust liability.

Case 6: Apple App Store / DMA Proceedings

The EU's Apple App Store proceedings provide another important example of adaptive competition regulation.

The DMA requires gatekeepers to allow developers to steer users toward alternative purchasing channels and establishes requirements concerning app distribution and access conditions.

In April 2025, the Commission found Apple in breach of the DMA's anti-steering obligation and imposed a €500 million fine.

Significance

The case illustrates the movement from:

"Was there an abuse of dominance?"

toward:

"What conduct must a systemically important ecosystem operator be prohibited from doing in advance?"

This is central to adaptive antitrust.

6. Additional Relevant Case: Google Search Advertising

In Google LLC and Alphabet Inc. v European Commission, Case T-334/19, the General Court examined Google's exclusive-supply obligations in the online search-advertising intermediation market.

The case demonstrates another ecosystem mechanism:

exclusive contractual arrangements → restricted rival access → reduced competitive opportunities.

In an autonomous ecosystem, similar restrictions could potentially be implemented through automated API permissions, developer terms, ranking mechanisms or machine-generated access decisions.

7. Competition Risks in Autonomous Ecosystems

A. Algorithmic self-preferencing

An AI platform may systematically rank:

• its own AI model;
• its own payment service;
• its own cloud service;
• its own marketplace;
• its own applications

above competing products.

The competitive concern is amplified when the algorithm continuously changes its ranking rules.

B. Automated exclusion

An ecosystem could automatically:

• deny API access;
• reduce visibility;
• increase commissions;
• restrict interoperability;
• impose technical barriers;
• classify competitors as risky;
• restrict data access.

Such conduct may be difficult for conventional enforcement to detect because there may be no obvious human instruction corresponding to every individual exclusion.

C. Algorithmic collusion

Several competitors may use a common algorithm or algorithmic intermediary.

Potential risks include:

Competitor data → common system → pricing recommendation → parallel implementation → reduced competition

The RealPage allegations demonstrate why this problem has become an important antitrust issue.

8. Autonomous Agents and Competition Law

The next stage involves AI agents acting on behalf of businesses or consumers.

For example:

AI procurement agents → automatically compare suppliers → negotiate prices → select suppliers → execute contracts.

Potential competition questions include:

• Can agents coordinate without explicit human communication?
• Who is responsible for an agent's anti-competitive conduct?
• Can competitors deploy interoperable agents?
• Can dominant platforms prevent rival agents from accessing data?
• Can an AI agent discriminate between suppliers?
• Can an ecosystem manipulate agents through rankings or default settings?

Competition law may therefore have to regulate not merely companies but the architecture through which autonomous agents interact.

9. Interoperability as an Antitrust Remedy

Interoperability may become one of the most important remedies.

A dominant autonomous ecosystem may need to permit rivals to interact with:

• operating systems;
• APIs;
• data;
• identity systems;
• payment systems;
• AI functionality;
• communication protocols.

The European Commission's July 2026 measures concerning Google provide a contemporary example: the Commission issued binding specifications intended to give competing AI services equal access to certain Android functionalities and to provide third-party search engines access to Google Search data.

This represents a movement toward competition by design.

10. Data as an Autonomous-Ecosystem Bottleneck

Data can simultaneously be:

• an input;
• a competitive advantage;
• a feedback mechanism;
• a training resource;
• an interoperability resource.

The competition problem becomes particularly serious where:

more users → more data → better AI → better service → more users → still more data.

This can create a self-reinforcing network effect.

Adaptive antitrust may therefore examine:

• data portability;
• data access;
• data combination;
• exclusive data arrangements;
• interoperability;
• data-sharing obligations.

11. Merger Control and Autonomous Ecosystems

Traditional merger thresholds may fail to capture acquisitions of emerging AI or technology companies whose current revenues are relatively low but whose future competitive importance is substantial.

Authorities may therefore examine:

• nascent competitors;
• AI startups;
• data acquisitions;
• talent acquisitions;
• interoperability assets;
• foundation models;
• cloud infrastructure;
• strategic APIs.

The relevant question becomes not merely:

"What is the target's current market share?"

but also:

"What competitive constraint might disappear if this technology becomes part of the dominant ecosystem?"

12. Adaptive Antitrust Enforcement Model

A useful framework can be represented as:

Market Monitoring
↓
Data Collection
↓
Algorithmic Audit
↓
Market-Power Assessment
↓
Risk Detection
↓
Investigation
↓
Interim Measures
↓
Behavioural / Interoperability Remedy
↓
Continuous Monitoring
↓
Remedy Adjustment

The critical feature is the feedback loop.

Unlike a traditional enforcement process that ends with a decision, an adaptive system continuously asks whether the remedy is still effective.

13. Ex Post vs Adaptive Antitrust

14. Due Process Challenges

Adaptive antitrust should not mean unrestricted regulatory discretion.

Important safeguards include:

Transparency

Businesses should understand the legal standards being applied.

Explainability

Where algorithmic evidence is relied upon, parties should have meaningful opportunities to challenge it.

Reproducibility

Technical findings should be capable of independent verification where possible.

Confidentiality

Competition authorities must protect trade secrets and sensitive data.

Human oversight

Automated regulatory systems should support, rather than replace, legally accountable decision-makers.

Proportionality

Remedies should correspond to the identified competitive harm.

15. Role of Competition Authorities

Future authorities may require multidisciplinary teams consisting of:

• competition lawyers;
• economists;
• AI specialists;
• data scientists;
• cybersecurity experts;
• software engineers;
• consumer researchers.

A competition authority investigating an autonomous ecosystem may need to inspect not only contracts and financial records but also:

• source-code documentation;
• API architecture;
• model documentation;
• audit trails;
• system logs;
• training-data practices;
• ranking criteria;
• automated decision rules.

16. Future Legal Doctrine

Several doctrines may evolve.

1. Dynamic dominance

Dominance may be assessed partly by an undertaking's ability to control the evolution of an ecosystem.

2. Algorithmic foreclosure

Exclusion may occur through automated ranking, access or pricing systems.

3. Ecosystem leverage

Power in one market may be leveraged into adjacent technological markets.

4. Data-based essentiality

Certain datasets may become competitively indispensable.

5. Interoperability duties

Dominant ecosystem operators may face stronger obligations to permit technically meaningful interoperability.

6. Continuous remedies

Remedies could be monitored and modified as market conditions change.

17. Application to India

For India, the principal framework is the Competition Act, 2002, administered by the Competition Commission of India.

The most relevant provisions include:

• Section 3 — anti-competitive agreements;
• Section 4 — abuse of dominant position;
• Section 5 — combinations;
• Section 19 — inquiry into agreements and dominance;
• Section 26 — investigation procedure;
• Section 27 — orders after finding contravention;
• Section 33 — interim orders.

For autonomous ecosystems, Section 4 is particularly relevant to:

• self-preferencing;
• discriminatory access;
• tying;
• leveraging;
• denial of market access;
• exclusionary ecosystem practices.

Section 3 becomes relevant to algorithmic coordination and information-sharing arrangements.

Combination control becomes important where large technology firms acquire:

• AI startups;
• data businesses;
• cloud infrastructure;
• emerging competitors;
• complementary ecosystem technologies.

18. Key Legal Lessons from the Case Laws

The Google Android judgment is particularly significant because the General Court expressly addressed the interaction between a multi-sided platform and an ecosystem.

19. Conclusion

Adaptive antitrust systems for autonomous ecosystems represent a transition from static competition enforcement to continuous ecosystem governance.

The central legal challenge is that future competitive conduct may no longer take the form of a single identifiable human decision. It may emerge from:

data + algorithms + AI + network effects + interoperability + automated contracts + ecosystem control.

The existing cases already provide important building blocks. Google Android demonstrates ecosystem-based analysis; Google Shopping illustrates self-preferencing; Qualcomm demonstrates exclusionary and pricing analysis; RealPage illustrates algorithmic coordination; and the DMA's Apple and Google proceedings demonstrate the increasing importance of ex ante, continuously monitored obligations.

The future of competition law is therefore likely to involve continuous monitoring, algorithmic auditing, interoperability requirements, data-access rules, dynamic merger scrutiny and technologically informed remedies, while preserving due process and human legal accountability.

Core proposition:

Where markets become autonomous, antitrust must become sufficiently adaptive to regulate not only market conduct, but also the technological architecture through which competitive conditions are continuously created and changed.