1. Introduction

Essay scoring engines are artificial intelligence (AI)-based systems designed to evaluate written assignments, examinations, academic essays and other forms of written communication. These systems use natural language processing (NLP), machine learning, predictive analytics and automated assessment models to assign marks, generate feedback and assess writing quality.

Examples of functions performed by essay scoring engines include:

Automated evaluation of grammar, vocabulary and sentence structure.

Assessment of coherence, originality and logical reasoning.

Detection of plagiarism and AI-generated content.

Automated grading of school, university and competitive examination essays.

Generation of performance reports and improvement recommendations.

Integration with learning management systems (LMS).

Provision of institutional assessment and examination services.

Essay scoring engines create significant competition-law concerns because a small number of technology companies may control the underlying algorithms, training datasets, institutional contracts, assessment infrastructure and distribution channels.

Competition law seeks to ensure that these technologies develop through innovation, quality, accuracy and competitive pricing rather than exclusionary conduct, collusion or exploitation of market power.

Central legal question: Whether companies developing and supplying automated essay scoring technologies use their technological advantages to improve educational assessment or to restrict competition and exclude competing assessment providers.

2. Nature of the Relevant Market

Competition authorities must identify the relevant product and geographical markets before examining alleged anticompetitive conduct.

A. Relevant product market

Possible product markets include:

Automated essay scoring software.

AI-based examination evaluation platforms.

Educational assessment and grading systems.

Plagiarism detection and originality assessment services.

Institutional assessment management software.

AI-powered writing evaluation and feedback platforms.

These products may constitute separate markets or overlapping segments depending on their functionality and substitutability.

For example, a university requiring automated evaluation of 100,000 examination essays may not consider ordinary grammar-checking software an effective substitute for a sophisticated scoring engine.

B. Relevant geographical market

The geographical market may be:

National, where educational procurement and regulatory requirements are country-specific.

Regional, where institutions use common assessment standards.

International, where cloud-based scoring engines compete across borders.

Data localisation, language requirements, accreditation standards and public procurement conditions may restrict geographical competition.

3. Major Competition-Law Issues

A. Abuse of dominant position

A company controlling a widely adopted essay scoring engine may acquire substantial market power through:

Exclusive institutional contracts.

Control over proprietary assessment datasets.

High switching costs.

Technological superiority and network effects.

Control over assessment APIs.

Long-term integration agreements with universities.

A dominant provider may violate competition law if it uses its position to eliminate competition rather than compete on the merits.

Under Indian competition law, Section 4 of the Competition Act, 2002 prohibits abuse of dominant position.

Potential abusive practices include:

Charging excessive institutional licensing fees.

Refusing access to essential interoperability facilities.

Imposing unfair contractual conditions.

Restricting competing scoring engines.

Bundling scoring engines with unrelated educational software.

Discriminating between similarly situated educational institutions.

Illustration: A dominant essay scoring provider refuses to provide API compatibility to universities that simultaneously purchase competing plagiarism detection software. Such conduct may raise concerns about exclusionary abuse.

B. Predatory pricing

Predatory pricing arises when a dominant company sacrifices profits through below-cost pricing to eliminate competitors and subsequently exploit its strengthened market position.

In essay scoring markets, a dominant provider might:

Offer free automated grading to universities.

Subsidise institutional subscriptions using revenue from other services.

Provide heavily discounted assessment contracts.

Sell scoring engines below sustainable cost.

Subsequently increase prices after competitors exit.

However, low prices are not automatically unlawful.

Competition authorities must examine the provider's market position, pricing methodology, exclusionary strategy and possibility of recoupment, where relevant under the applicable legal standard.

C. Exclusive agreements with universities

Exclusive agreements may prevent universities and examination boards from purchasing competing scoring technologies.

Potential restrictions include:

Exclusive institutional licensing.

Long-term procurement agreements.

Minimum purchase commitments.

Restrictions on parallel testing of alternative engines.

Contractual prohibitions on using competing assessment systems.

Such agreements may foreclose competing developers from accessing important customers.

In India, Section 3(4) of the Competition Act addresses specified vertical agreements, including exclusive supply, exclusive distribution, tie-in arrangements and refusal to deal, where the statutory conditions are satisfied.

D. Algorithmic collusion

Algorithmic collusion is an important concern in automated essay scoring markets.

Competing providers may use pricing algorithms that independently or jointly facilitate coordination.

Examples include:

Similar automated subscription pricing.

Coordinated institutional licensing fees.

Exchange of competitively sensitive procurement information.

Common pricing software used by competing providers.

Algorithmic monitoring of competitors' bids.

The mere use of similar algorithms or parallel pricing does not establish an unlawful agreement.

Authorities must establish the necessary agreement, coordination or legally relevant concerted conduct.

E. Data monopolisation

Training datasets are important competitive inputs for essay scoring engines.

A company with access to extensive collections of evaluated essays, examination responses and human-assigned grades may develop more sophisticated scoring models.

Competition concerns arise where a dominant provider:

Acquires exclusive rights over large educational datasets.

Prevents institutions from transferring their assessment data.

Restricts legitimate portability of historical scoring records.

Denies competitors access to indispensable data without objective justification.

Uses institutionally generated data to strengthen its own competing educational services.

Data accumulation alone does not constitute an infringement. Its competitive significance depends on exclusivity, replicability, substitutability and the actual foreclosure effects.

F. Self-preferencing and discriminatory scoring

A company supplying both essay scoring technology and educational content may favour its own educational products.

For example, its scoring engine may systematically reward essay structures resembling its own proprietary learning materials.

Other possible concerns include:

Preferential treatment of affiliated educational platforms.

Discriminatory API access.

Artificially lower scoring accuracy for competing platforms.

Biased recommendations for educational resources.

Unfair ranking of competing assessment providers.

Authorities must distinguish genuine algorithmic bias or technical limitations from deliberate exclusionary conduct.

G. Interoperability restrictions

Interoperability allows universities to integrate different assessment technologies.

A dominant provider may restrict interoperability by:

Using proprietary file formats.

Withholding API documentation.

Restricting access to scoring outputs.

Preventing integration with competing LMS platforms.

Imposing unreasonable technical certification requirements.

Such practices may increase switching costs and create technological lock-in.

H. Merger and acquisition concerns

Competition authorities should carefully scrutinise acquisitions involving:

AI essay scoring startups.

Educational assessment platforms.

Plagiarism detection providers.

Academic data analytics businesses.

NLP technology developers.

Examination management companies.

A dominant educational technology provider acquiring an emerging scoring engine may eliminate a potential competitive constraint.

Merger assessment should consider innovation competition, access to training data, entry barriers and future technological developments.

4. Legal Framework in India

The principal statutory framework is the Competition Act, 2002, as amended.

The Competition Commission of India (CCI) may examine whether an essay scoring provider possesses market power and whether its conduct produces exclusionary or exploitative effects.

Section 3(3) requires particular attention where competing assessment companies coordinate prices, tenders or customer allocation.

Section 4 is particularly relevant where a dominant scoring platform uses its control over technology, data or institutional access to exclude competitors.

5. Important Case Laws

Important clarification: There is no established body of reported appellate judgments specifically concerning AI-based essay scoring engines. The following cases provide important analogous principles from digital platforms, algorithmic coordination, market foreclosure and competition enforcement.

1. Samir Agrawal v. Competition Commission of India (2021) 3 SCC 136

Court: Supreme Court of India.

Legal principle: Algorithmic technology and pricing arrangements must be examined through the requirements of competition law.

The case concerned allegations of price coordination involving cab aggregators and their drivers through platform technology.

The Supreme Court considered whether the alleged arrangement satisfied the requirements of an anticompetitive agreement.

Application to essay scoring engines:

Common algorithms do not automatically establish collusion.

The relationship between technology providers and users requires careful examination.

Authorities must establish the necessary elements of an anticompetitive agreement.

Automated pricing arrangements can attract competition scrutiny.

This judgment is particularly relevant to algorithmically coordinated pricing among competing essay scoring providers.

2. Excel Crop Care Ltd. v. Competition Commission of India (2017) 8 SCC 47

Court: Supreme Court of India.

Legal principle: Bid rigging and coordinated conduct can undermine competitive procurement.

The case concerned collusive bidding in a public procurement process involving aluminium phosphide tablets.

The Supreme Court examined the competition implications of coordinated tender participation and the determination of penalties.

Application to essay scoring engines:

Suppose several essay scoring companies coordinate their bids for a national examination board's automated evaluation contract.

Possible violations include:

Bid rotation.

Artificially inflated tender prices.

Allocation of examination boards.

Suppression of competitive bidding.

Exchange of confidential tender information.

The case demonstrates why public procurement of educational assessment technology must remain competitive.

3. Rajasthan Cylinders and Containers Ltd. v. Union of India (2018) 13 SCC 472

Court: Supreme Court of India.

Legal principle: Parallel commercial behaviour alone is insufficient to establish cartelisation.

The case concerned allegations of cartelisation in the supply of LPG cylinders.

The Supreme Court considered market conditions and the distinction between parallel conduct and collusive conduct.

Application to essay scoring engines:

If several AI scoring providers introduce similar subscription prices, this does not automatically establish price fixing.

Authorities should examine:

Communications between competitors.

Exchange of confidential pricing information.

Common commercial arrangements.

Market conditions.

Evidence of coordination.

The judgment protects legitimate independent commercial decision-making while permitting intervention where collusion is established.

4. Competition Commission of India v. Steel Authority of India Ltd. (2010) 10 SCC 744

Court: Supreme Court of India.

Legal principle: The CCI's statutory investigation framework and procedural powers must operate within the Competition Act.

The Supreme Court considered the nature of directions for investigation under the competition-law framework.

Application to essay scoring engines:

The CCI may examine complaints concerning:

Exclusive scoring-engine contracts.

Discriminatory institutional licensing.

Refusal of API access.

Predatory pricing allegations.

Anticompetitive procurement arrangements.

The case is important for understanding the procedural foundation of competition enforcement in this sector.

5. Eturas UAB v. Lietuvos Respublikos konkurencijos taryba, Case C-74/14 (2016)

Court: Court of Justice of the European Union.

Legal principle: Restrictions communicated through a common electronic platform may contribute to an unlawful concerted practice, subject to the applicable evidentiary requirements.

The case concerned a centralised online booking system that communicated restrictions on discounts to participating travel agencies.

Application to essay scoring engines:

Imagine competing scoring companies using a shared commercial platform that communicates restrictions on discounts offered to universities.

The case is relevant to:

Common pricing infrastructure.

Electronic communication of commercial restrictions.

Platform-mediated coordination.

Exchange of competitively sensitive information.

It demonstrates that digital communication mechanisms do not provide immunity from competition law.

6. United States v. Topkins, No. CR 15-00201 (N.D. Cal. 2015)

Court: United States District Court for the Northern District of California.

Legal principle: Algorithmic tools can facilitate the implementation of an unlawful price-fixing agreement.

The case involved online sellers using pricing algorithms in connection with an agreement to coordinate prices for posters sold through an online marketplace.

Application to essay scoring engines:

Competing scoring providers might use a common algorithm to implement agreed institutional subscription prices.

Competition authorities should investigate:

Common pricing software.

Agreements concerning pricing parameters.

Communications between providers.

Coordinated adjustments to subscription fees.

Algorithmic implementation of an underlying agreement.

The case illustrates that automation does not remove liability for conventional cartel conduct.

7. United States v. Apple Inc., 791 F.3d 290 (2d Cir. 2015)

Court: United States Court of Appeals for the Second Circuit.

Legal principle: A platform intermediary may be liable for facilitating a horizontal price-fixing conspiracy.

The case concerned Apple's role in a conspiracy with major publishers to increase e-book prices through a new distribution arrangement.

Application to essay scoring engines:

A technology intermediary supplying a common assessment marketplace could potentially facilitate coordination between competing scoring providers.

For example, an intermediary might:

Coordinate institutional subscription prices.

Restrict discounts.

Allocate universities between providers.

Facilitate exchanges of competitively sensitive information.

The case is relevant to platform-facilitated collusion and intermediary responsibility.

8. Google LLC v. European Commission (Google Shopping), Case C-48/22 P (2024)

Court: Court of Justice of the European Union.

Legal principle: A dominant platform's preferential treatment of its own services may constitute an abuse of dominant position where the applicable exclusionary-effect requirements are satisfied.

The case concerned Google's preferential positioning of its own comparison-shopping service in general search results.

Application to essay scoring engines:

A dominant provider might operate both an essay scoring engine and an educational marketplace.

It could potentially favour its own scoring products by:

Giving its own engine preferential placement.

Restricting competitors' visibility.

Providing discriminatory access to assessment interfaces.

Using its control over institutional access to disadvantage rival scoring companies.

The judgment is important for analysing self-preferencing and exclusionary conduct in digital markets.

9. United States v. Google LLC, 2020

Court: United States District Court for the District of Columbia.

Legal principle: Exclusionary agreements by a dominant digital platform may raise monopolisation concerns under Section 2 of the Sherman Act.

The case concerned Google's distribution agreements and alleged maintenance of monopoly power in general search services.

Application to essay scoring engines:

A dominant scoring provider might enter into agreements with major universities, examination boards or LMS providers that make rival scoring technologies commercially inaccessible.

The case provides an important framework for assessing distribution restrictions, market access and exclusionary contractual arrangements.

6. Specific Competition Scenarios in Essay Scoring Markets

7. Role of Artificial Intelligence and Algorithmic Transparency

AI-powered scoring engines introduce competition concerns that traditional educational assessment systems may not have encountered.

A. Proprietary training datasets

Large datasets can create competitive advantages because the quality of training data may influence scoring accuracy.

However, authorities should distinguish legitimate intellectual property protection from exclusionary data practices.

B. Explainability and transparency

Where an engine's scoring criteria are completely opaque, universities may find it difficult to compare competing products.

Competition authorities may examine whether technical opacity is being deliberately used to create switching costs or prevent effective competition.

C. Algorithmic discrimination

A scoring engine may systematically disadvantage particular writing styles, language varieties or educational backgrounds.

Such conduct may raise consumer protection, education-law, equality and data-protection concerns in addition to competition issues.

Competition authorities should investigate whether discriminatory technical design also operates as an exclusionary commercial strategy.

D. Interoperability and portability

Educational institutions should be able to transfer relevant assessment records and integrate alternative scoring engines without unreasonable technical restrictions.

Appropriate interoperability can lower entry barriers and encourage innovation.

E. Innovation competition

Competition authorities should examine not only present market shares but also future competitive constraints.

A startup developing a more accurate essay scoring engine may represent a significant competitive threat even before acquiring substantial institutional customers.

8. Economic Analysis of Competition

The following economic indicators can help competition authorities assess the market.

1. Market concentration

The Herfindahl-Hirschman Index (HHI) can help evaluate market concentration among essay scoring providers.

2. Switching costs

Authorities should examine the financial and technical costs universities incur when changing assessment providers.

3. Network effects

A scoring engine may become more attractive as more institutions use it and generate compatible assessment data.

4. Economies of scale

Large providers may spread infrastructure and model-development costs across a larger customer base.

5. Barriers to entry

Important barriers include proprietary datasets, computing infrastructure, institutional accreditation, reputation and long-term contracts.

6. Price-cost analysis

Authorities may examine whether pricing reflects legitimate efficiency advantages or an exclusionary strategy.

7. Innovation analysis

The impact of the alleged conduct on the development of alternative scoring technologies should also be evaluated.

9. Remedies and Regulatory Measures

Competition authorities may consider the following remedies where an infringement is established.

A. Structural remedies

Divestiture in appropriate merger or dominance cases.

Separation of conflicting business operations where legally justified.

Prevention of anticompetitive acquisitions.

B. Behavioural remedies

Termination of exclusionary contractual clauses.

Non-discriminatory API access.

Fair and transparent licensing conditions.

Removal of unjustified exclusivity requirements.

Prohibition of coordinated pricing arrangements.

C. Data-related remedies

Appropriate data portability.

Non-discriminatory access to relevant datasets where legally justified.

Restrictions on misuse of confidential institutional information.

Technical standards supporting interoperability.

D. Institutional safeguards

Independent algorithmic audits.

Competition compliance programmes.

Transparent procurement procedures.

Monitoring of market concentration.

Periodic review of institutional licensing arrangements.

Remedies should be proportionate and should not unnecessarily undermine legitimate intellectual property rights, privacy, cybersecurity or technological innovation.

10. Challenges in Competition-Law Enforcement

Competition authorities face several difficulties when investigating essay scoring engines.

1. Defining the market: Automated essay scoring may overlap with general educational software, examination management and AI assessment markets.

2. Technical complexity: Authorities may struggle to understand proprietary scoring algorithms and their commercial effects.

3. Data dependency: It may be difficult to determine whether a dataset is genuinely indispensable to competitors.

4. Rapid technological development: New AI models can change market conditions quickly.

5. Cross-border operations: Providers may operate through international cloud infrastructure and foreign subsidiaries.

6. Distinguishing innovation from exclusion: Superior accuracy and efficient technology should not be treated as unlawful merely because competitors cannot match them.

7. Measuring competitive harm: Authorities must distinguish reduced competition from legitimate differences in product quality.

11. Practical Hypothetical Case Study

Facts:

A company called EduScore AI develops a widely used automated essay scoring engine.

It enters into agreements with 75% of universities in a particular national market.

The company subsequently:

Introduces exclusive five-year institutional agreements.

Refuses API access to competing educational platforms.

Offers below-cost subscriptions to new universities.

Prevents participating institutions from transferring historical scoring data.

Acquires a startup developing a competing multilingual scoring engine.

A competing provider files a complaint before the competition authority.

Legal issues:

Whether EduScore AI possesses a dominant position.

Whether exclusive agreements foreclose competing scoring providers.

Whether its pricing constitutes predatory pricing.

Whether restrictions on data portability constitute abusive conduct.

Whether API restrictions create unjustified interoperability barriers.

Whether acquisition of the multilingual startup substantially reduces competition.

Possible legal findings:

The authority should investigate each practice separately and collectively.

It must establish relevant markets, dominance, exclusionary effects and applicable statutory requirements. The acquisition should be examined under the relevant merger-control thresholds and jurisdictional rules.

If infringements are established, appropriate remedies may include removal of exclusionary clauses, non-discriminatory interoperability, proportionate penalties and merger remedies where legally available.

12. Conclusion

Competition law plays an important role in ensuring that essay scoring engines develop as competitive, innovative and accessible educational technologies.

The most significant competition concerns involve abuse of dominance, algorithmic collusion, predatory pricing, exclusive university agreements, data monopolisation, discriminatory access, self-preferencing and anticompetitive acquisitions.

The cases of Samir Agrawal, Excel Crop Care, Eturas, Topkins, United States v. Apple and Google Shopping provide useful legal principles for analysing these emerging issues.

The central objective should be to preserve competition in educational AI markets without unnecessarily restricting technological innovation.