Competition Law and Autonomous Education Ecosystems

1. Introduction

An autonomous education ecosystem is a digitally integrated education environment in which software, artificial intelligence (AI), learning-management systems, adaptive-learning engines, assessment tools, digital content, student-data systems, payment services, and institutional platforms interact with one another and increasingly make decisions with limited human intervention.

Examples include:

• AI tutors that independently personalize learning;
• adaptive examination and assessment platforms;
• AI-generated educational content;
• automated student-placement and recommendation systems;
• school-management ecosystems;
• learning-management systems integrated with app stores and cloud services;
• platforms controlling teachers, students, content providers and educational institutions;
• AI systems using student-performance data to determine course recommendations.

Competition law becomes important when one undertaking controls several interconnected layers of this ecosystem. The concern is not simply that an education platform becomes large. The issue is whether ecosystem power is used to exclude competitors, restrict interoperability, exploit dependent users, or extend dominance from one educational market into another.

In China, this analysis is particularly relevant because the Anti-Monopoly Law (AML), the Platform Economy Anti-Monopoly Guidelines and digital-platform enforcement provide tools for addressing platform dominance, exclusivity, tying, discriminatory treatment and data-related competitive effects. China's regulatory framework also intersects with education regulation, particularly for online education and compulsory-education services.

2. Meaning of an Autonomous Education Ecosystem

An autonomous education ecosystem can be represented as:

AI/Algorithmic Core
↓
Learning Platform / LMS
↓
Student Data + Learning Records
↓
Content + Assessment + Tutoring
↓
Schools / Universities / Teachers
↓
Students / Parents
↓
Payments + Certification + Employment Services

The important competition-law feature is interdependence.

A platform may simultaneously operate:

1. the learning-management system;
2. the AI tutor;
3. the assessment system;
4. the educational-content marketplace;
5. the student-data infrastructure;
6. the cloud infrastructure;
7. the payment system; and
8. the application marketplace.

This creates opportunities for vertical leveraging.

For example:

A dominant LMS could make access to its AI tutor conditional on using its own assessment system and then make competing assessment providers technically incompatible.

That conduct could potentially raise questions of tying, refusal of access, interoperability restrictions, self-preferencing and exclusionary abuse.

3. Relevant Competition-Law Framework in China

A. Anti-Monopoly Law

The principal legal framework is China's Anti-Monopoly Law, particularly rules dealing with:

• monopoly agreements;
• abuse of dominant market position;
• concentrations of undertakings;
• digital-platform conduct;
• discriminatory treatment;
• tying;
• unreasonable trading conditions;
• refusal to deal; and
• exclusionary conduct.

The 2022 amendments strengthened the framework applicable to digital-platform markets.

B. Platform Economy Guidelines

China's Anti-Monopoly Guidelines for the Platform Economy Sector are particularly relevant.

An autonomous education ecosystem can constitute a multi-sided platform involving:

• students;
• parents;
• schools;
• teachers;
• content providers;
• publishers;
• examination providers;
• advertisers; and
• employers.

The competition analysis therefore cannot necessarily be confined to the price paid by students.

4. Relevant Markets

A regulator may examine several separate but interconnected markets.

4.1 Online education services

This may include:

• online tutoring;
• professional education;
• university education;
• examination preparation;
• language learning; and
• vocational education.

4.2 Learning-management systems

The relevant market may consist of software supplied to:

• schools;
• universities;
• corporations; or
• individual learners.

4.3 AI educational services

A separate market may potentially emerge for:

• AI tutoring;
• automated assessment;
• AI content generation;
• personalized learning;
• educational recommendation engines.

4.4 Educational data services

Student-performance data can become an important competitive input.

4.5 Educational app ecosystems

Where one platform controls the operating environment through which educational applications reach users, competition between educational applications may depend upon access to that ecosystem.

5. Ecosystem Market Power

Traditional competition analysis often asks:

"What is the relevant product market?"

Autonomous ecosystems require an additional question:

"Where does the undertaking's ecosystem create competitive dependence?"

A company might have only moderate market share in AI tutoring but possess enormous leverage because it controls:

• the school's LMS;
• student identity;
• examination records;
• cloud infrastructure;
• payment facilities; and
• app distribution.

This is sometimes described as ecosystem power.

The European Commission's digital-market framework expressly recognizes the significance of platforms functioning as gateways, while recent EU jurisprudence concerning Google's Android ecosystem illustrates how several interconnected products can be assessed together.

6. Major Competition Concerns

6.1 Self-Preferencing

A dominant education ecosystem may rank its own:

• AI tutor;
• digital textbooks;
• examination platform;
• educational videos; or
• certification services

above competing products.

For example, an LMS could systematically place its own AI tutor at the top of search results while reducing the visibility of independent tutors.

The concern becomes stronger where the platform is an unavoidable gateway to students.

The Google Android litigation is important by analogy because the EU courts examined the relationship between an operating system, app store, search engine and browser, including contractual restrictions and exclusionary effects.

7. Tying and Bundling

An autonomous education provider could require schools purchasing its LMS to also purchase:

• its AI tutor;
• its assessment system;
• its cloud storage;
• its educational content; or
• its payment service.

Example:

School purchases LMS → AI examination system automatically included → competing examination providers cannot be installed.

Possible competition theories include:

• tying;
• bundling;
• foreclosure;
• leveraging;
• exclusion of rival suppliers.

The Google Android case is again relevant because the European litigation concerned contractual restrictions and tying across interconnected digital products.

8. Interoperability Restrictions

This may be one of the most important issues.

Suppose a dominant education platform refuses to permit competing AI systems to connect with:

• student records;
• learning histories;
• examination databases;
• digital classrooms; or
• identity-management systems.

The resulting switching costs may become substantial.

Students and schools may remain with the dominant platform because transferring several years of learning data is expensive or technically impossible.

Potential concerns include:

• refusal to deal;
• exclusionary conduct;
• interoperability foreclosure;
• discriminatory API access; and
• creation of artificial switching costs.

9. Data-Driven Market Power

Autonomous education platforms may accumulate enormous quantities of:

• student performance data;
• learning histories;
• examination results;
• behavioural data;
• teacher evaluations;
• attendance records;
• course preferences.

The competitive advantage comes from the feedback loop:

More students → more data → better AI → better personalization → more schools → more students → still more data

This can produce strong network effects.

A new competitor may therefore face a structural disadvantage even if its AI technology is technically comparable.

Competition analysis should distinguish between:

• lawful economies of scale;
• legitimate data advantages; and
• exclusionary acquisition or use of data.

10. Algorithmic Discrimination

Autonomous systems can make competitive decisions automatically.

For example, an algorithm may determine:

• which educational provider appears first;
• which tutor receives students;
• which courses are recommended;
• which advertisements appear;
• which educational content receives distribution;
• which schools receive preferential pricing.

If the algorithm systematically disadvantages competitors controlled by independent suppliers, the conduct may raise competition concerns.

The fact that the decision is made by an algorithm does not automatically remove the undertaking's responsibility for the competitive consequences.

11. Algorithmic Collusion

Two autonomous education platforms could use pricing algorithms to monitor:

• tuition fees;
• tutoring prices;
• examination fees;
• subscription prices.

If competitors deliberately design systems to coordinate prices or stabilize market outcomes, traditional rules against concerted practices or monopoly agreements may become relevant.

A distinction must be maintained between:

independent algorithmic adaptation, which is not automatically unlawful, and

intentional coordination between competitors, which can raise serious antitrust concerns.

12. Exclusive Contracts with Schools

A dominant education platform might require schools to:

• use only its LMS;
• purchase only its AI tutor;
• prohibit competing educational applications;
• give it exclusive access to student data;
• prohibit teachers from using competing platforms.

Exclusivity can be especially significant where the platform controls a large proportion of schools.

The competitive question is whether rivals are effectively prevented from obtaining sufficient scale to compete.

13. Lock-In and Switching Costs

Autonomous education ecosystems can create unusually strong lock-in because the platform may contain:

• years of student records;
• personalized AI models;
• teacher materials;
• examination databases;
• school administrative records;
• learning analytics.

A school might technically be able to leave the platform but practically find migration prohibitively expensive.

Competition law may therefore examine effective switching barriers, not merely contractual exit rights.

14. Access to Educational APIs

APIs can become essential infrastructure.

A dominant platform may control APIs connecting:

School → LMS → AI Tutor → Assessment → Student Record

If the platform provides its own AI tutor unrestricted API access but gives competitors:

• delayed access;
• inferior technical access;
• incomplete data;
• discriminatory pricing; or
• no access,

this could raise discrimination and foreclosure concerns.

15. Essential-Facility Issues

The strongest cases may arise where the educational platform controls infrastructure that competitors cannot reasonably reproduce.

Potential examples include:

• national-scale examination infrastructure;
• uniquely large educational databases;
• dominant school identity systems;
• indispensable digital certification infrastructure.

However, being important is not automatically enough to constitute an essential facility.

The legal analysis generally requires careful examination of:

1. indispensability;
2. absence of realistic alternatives;
3. ability to reproduce the facility;
4. dominance;
5. refusal or discriminatory access;
6. competitive foreclosure; and
7. objective justification.

16. Merger Control

Autonomous education ecosystems also create significant merger concerns.

A large platform acquiring:

• an AI tutor;
• an examination company;
• a student-data provider;
• a digital textbook publisher;
• a school-management platform

could eliminate an emerging competitive constraint.

Traditional turnover thresholds may not always capture the strategic significance of small but technologically important targets.

China's enforcement against TAL Education Group/DaDa Education demonstrates that an education-technology transaction can trigger merger-control obligations. SAMR penalized TAL in March 2021 for implementing the acquisition of a majority stake in DaDa without prior notification.

17. Case Laws

The following cases are particularly useful for analysing autonomous education ecosystems. Some are direct education cases, while others are digital-platform cases whose principles apply by analogy.

Case 1 — TAL Education Group / DaDa Education (China, 2021)

Issue: Failure to notify a concentration involving an online education platform.

TAL acquired a majority interest in DaDa Education. SAMR found that the transaction had been implemented without the required prior notification and imposed a penalty.

Principle:

Digital education transactions are not outside merger control merely because the underlying service is educational.

Relevance to autonomous education ecosystems:

An acquisition of an AI-learning platform, adaptive-testing provider or student-data platform may need to be examined before completion.

Case 2 — Beijing Education and Training Enterprise RPM Case (China)

Chinese enforcement has also addressed resale-price-maintenance restrictions in the education and training sector. The case is significant because it concerned a commercial-franchise model and demonstrated that vertical restrictions imposed by an education franchisor can still attract competition-law scrutiny.

Principle:

Commercial justification does not automatically immunize contractual restrictions from antitrust scrutiny.

Relevance:

An autonomous education ecosystem using franchise schools could attempt to impose:

• mandatory subscription prices;
• fixed course prices;
• minimum prices for AI tutoring;
• restrictions on competing educational products.

Such restrictions must be examined under the applicable vertical-agreement rules.

Case 3 — Alibaba / "Choose One from Two" (China, 2021)

SAMR imposed a major penalty on Alibaba for its exclusive dealing practice commonly described as "choose one from two."

The case is important for autonomous education ecosystems because the same structural problem can arise where a dominant educational platform requires schools, teachers or educational suppliers to deal exclusively with it.

Principle:

A dominant platform cannot use ecosystem power to systematically restrict business partners from dealing with competing platforms.

Education application:

A dominant education marketplace might tell schools:

"Use our AI learning platform, or you cannot access our student marketplace."

That could raise analogous exclusionary concerns.

Case 4 — Google Android (Google and Alphabet v European Commission, T-604/18; C-738/22 P)

This is one of the most important ecosystem cases.

The EU proceedings concerned Google's Android ecosystem, including:

• Android;
• Google Search;
• Chrome;
• Google Play;
• device manufacturers;
• contractual restrictions;
• tying;
• exclusivity payments; and
• anti-fragmentation obligations.

The General Court expressly discussed the concept of a multi-sided platform and ecosystem. The Court of Justice issued its appeal judgment on 2 July 2026.

Relevance:

An education ecosystem could similarly combine:

operating environment + LMS + AI tutor + app marketplace + search/recommendation + data.

The case demonstrates why competition authorities may examine the interaction between different products rather than treating every component in isolation.

Case 5 — United States v. Pearson plc / Reed Elsevier–Harcourt (2008)

This is a directly relevant education-sector merger case.

The U.S. Department of Justice challenged aspects of the Pearson/Reed Elsevier acquisition involving educational and assessment-related businesses. The matter ultimately resulted in a final judgment with structural remedies.

Principle:

Competition authorities can scrutinize concentration involving educational content, assessment and related educational-support markets.

Relevance:

An autonomous education ecosystem combining:

• textbooks;
• testing;
• assessment;
• AI tutoring; and
• educational databases

could create horizontal and vertical competitive concerns.

Case 6 — Microsoft/Activision Blizzard

Although not an education case, this merger is significant for ecosystem theory.

The transaction raised questions about whether a powerful technology ecosystem could use control of an important product to strengthen its position in related markets.

The broader significance is the recognition that competition analysis may need to consider:

• ecosystem effects;
• future competition;
• access to important inputs;
• foreclosure;
• dynamic competition.

Recent scholarship comparing Microsoft/Activision with Booking/eTraveli identifies precisely this tension between conventional market definition and ecosystem-level assessment.

Education application:

An education-platform operator acquiring a major AI tutor should not necessarily be assessed only by today's market shares. The authority may need to consider how the acquisition changes future ecosystem competition.

Case 7 — Google Search / Search Distribution Litigation

The U.S. Google litigation concerning search distribution is another important analogy.

The central competition question includes the use of agreements and distribution arrangements to preserve a powerful position in search.

Education application:

Suppose an education operating system automatically sets its own AI tutor as the default learning assistant.

The relevant question becomes whether default placement and distribution arrangements deprive rival AI tutors of effective access to users.

The Google litigation remains active in remedies and compliance proceedings as of September 2026.

Case 8 — Google Search Self-Preferencing under the EU Digital Markets Act

The EU's Digital Markets Act provides an especially relevant modern framework for ecosystem conduct.

In July 2026, the European Commission fined Google for DMA non-compliance involving self-preferencing in Search and restrictions on steering users toward alternative purchasing channels on Google Play.

Education application:

An educational gatekeeper could potentially face analogous concerns if it:

• ranks its own AI tutor above competitors;
• gives its own courses superior placement;
• restricts links to competing educational services; or
• prevents schools from directing users to alternative platforms.

18. Six Core Competition Theories

19. Autonomous Decision-Making and Attribution of Liability

A particularly difficult issue is:

Who is legally responsible when the algorithm makes the decision?

The answer should not simply be:

"The AI did it."

Competition authorities may investigate:

1. who designed the algorithm;
2. what objectives were programmed;
3. what data were supplied;
4. whether management knew about the conduct;
5. whether discriminatory outcomes were foreseeable;
6. whether the company monitored the system;
7. whether safeguards existed; and
8. whether the company benefited from the exclusionary effects.

Thus, algorithmic autonomy does not necessarily equal legal autonomy.

20. Network Effects

Education platforms can experience powerful direct and indirect network effects.

Direct effect

More students make the platform more attractive to teachers.

Indirect effect

More students → more teachers → more educational content → more students.

Data effect

More students → more learning data → better AI → better personalization → more students.

The third feedback loop can be particularly powerful in autonomous education systems.

21. Consumer Welfare and Educational Quality

Competition law in education has a broader practical dimension because the "consumer" may be:

• a student;
• a parent;
• a school;
• a university;
• a teacher; or
• a government education department.

Competitive harm may therefore involve more than higher prices.

Potential effects include:

• reduced educational choice;
• reduced innovation;
• inferior AI quality;
• reduced privacy or data control;
• weaker interoperability;
• discriminatory recommendations;
• reduced teacher autonomy;
• reduced availability of alternative educational content.

22. Objective Justification

Not every restrictive feature is unlawful.

An education platform may legitimately restrict interoperability where the restriction is necessary for:

• cybersecurity;
• child safety;
• examination integrity;
• protection against fraud;
• data security;
• technical reliability;
• intellectual-property protection.

The important question is whether the restriction is necessary, proportionate and objectively justified, rather than simply serving to protect the platform from competition.

23. Possible Remedies

Competition authorities could potentially consider:

Structural remedies

• divestiture;
• separation of AI and LMS businesses;
• separation of marketplace and educational services.

Behavioural remedies

• non-discrimination;
• interoperability;
• API access;
• data portability;
• transparent ranking;
• prohibition of self-preferencing;
• prohibition of exclusive dealing.

Merger remedies

• divestiture of competing products;
• licensing of educational content;
• access commitments;
• data-access commitments;
• interoperability commitments.

24. Compliance Framework for Education Platforms

An autonomous education platform should establish:

1. Market-power assessment

Identify markets in which the platform has significant market power.

2. Ecosystem mapping

Map:

LMS → AI → Data → Content → Assessment → Payments → Students.

3. Algorithm audit

Test whether algorithms systematically disadvantage rivals.

4. API policy

Ensure objective and non-discriminatory interoperability rules.

5. Data portability

Allow lawful transfer of educational records.

6. Ranking transparency

Document the factors used to rank educational services.

7. Exclusivity review

Review school, teacher and content-provider contracts.

8. Merger review

Examine acquisitions of small AI, data or educational-technology companies even where their current revenues are modest.

9. Internal antitrust controls

Train product and engineering teams—not merely legal teams—because competition risks can arise from technical design choices.

25. Special Importance of China

China presents a particularly interesting environment because education platforms sit at the intersection of:

Education Regulation + Digital Platform Regulation + Competition Law + Data Regulation + AI Governance.

Chinese regulators have already demonstrated that online education platforms can attract regulatory intervention. In 2021, for example, Yuanfudao and Zuoyebang were fined for misleading claims concerning their educational businesses and related practices.

At the same time, China's platform-antitrust enforcement has developed substantially beyond traditional price competition, including scrutiny of exclusivity and platform leverage.

This means an autonomous education ecosystem should not be analysed solely under traditional education regulation.

26. Hypothetical Example

Assume EduAI operates:

• China's largest school LMS;
• an AI tutor;
• an examination platform;
• a digital textbook marketplace;
• a teacher marketplace.

EduAI requires every school using its LMS to:

1. use EduAI's AI tutor;
2. use EduAI's examination system;
3. give EduAI exclusive access to learning data;
4. prohibit competing AI tutors;
5. place EduAI's courses first in search results.

The competition analysis could be:

LMS dominance
↓
Control over schools
↓
Control over student data
↓
AI-tutor leveraging
↓
Exclusion of rival AI providers
↓
Higher switching costs
↓
Reduced innovation and choice

Possible legal theories could include:

• tying;
• exclusive dealing;
• self-preferencing;
• discriminatory access;
• refusal/interoperability restrictions;
• abuse of dominance; and
• potentially merger-control concerns if the ecosystem was created through acquisitions.

27. Key Legal Principles from the Cases

28. Conclusion

Autonomous education ecosystems represent a shift from competition between individual educational products to competition between interconnected technological ecosystems.

The principal competition-law risks are:

1. ecosystem leveraging;
2. self-preferencing;
3. AI-tutor tying and bundling;
4. exclusive school contracts;
5. API and interoperability restrictions;
6. student-data foreclosure;
7. algorithmic discrimination;
8. algorithmic coordination;
9. student and institutional lock-in;
10. acquisition of emerging competitors; and
11. control of essential educational infrastructure.

The central legal question is therefore not merely whether an autonomous education platform is large. It is whether control over one layer of the education ecosystem is being used to restrict competition in another layer.