Competition Law and Competition Governance in Automated Public Systems

Introduction

Automated public systems are public-sector or publicly supervised systems in which algorithms, artificial intelligence, data-driven models, automated procurement platforms, digital identity systems, allocation mechanisms, regulatory technology, smart infrastructure, or automated decision-making tools substantially determine how public resources, contracts, licences, services, or opportunities are allocated.

Examples include:

• automated government procurement platforms;
• algorithmic allocation of public contracts;
• automated tax and customs systems;
• digital public-service platforms;
• automated licensing and permitting;
• public-sector cloud and AI procurement;
• smart-grid and transport allocation systems;
• algorithmic distribution of subsidies and benefits;
• automated public housing or healthcare allocation;
• government-operated digital marketplaces.

Competition law becomes relevant because automation can increase transparency and efficiency, but it can also create new forms of market power. A public authority may simultaneously act as regulator, purchaser, platform operator, data controller and market participant. This creates competition risks that traditional competition law was not designed to address fully.

I. Meaning of Competition Governance in Automated Public Systems

Competition governance means the institutional and legal framework through which competition is:

1. protected;
2. monitored;
3. facilitated;
4. corrected when distorted; and
5. reconciled with legitimate public objectives.

In automated public systems, governance must address not merely the conduct of human firms but also the architecture and operation of automated decision systems.

A useful conceptual model is:

Public authority → digital infrastructure → algorithmic decision → market allocation → competitive effects → regulatory oversight

Thus, competition law must examine both market behaviour and the design of the automated system.

II. Why Automated Public Systems Create Competition Concerns

1. Algorithmic discrimination

An automated procurement or licensing system may systematically favour:

• incumbent suppliers;
• firms with larger datasets;
• firms already integrated into government infrastructure;
• preferred technological standards;
• particular platforms or vendors.

Even without an express discriminatory instruction, algorithmic design can produce discriminatory competitive effects.

2. Preferential access to public data

Government systems frequently possess exceptionally valuable datasets.

Examples include:

• transport data;
• health data;
• procurement data;
• land records;
• business registrations;
• energy consumption;
• taxation information.

If a public authority gives its own commercial affiliate or a preferred private contractor superior access to such data, competitors may face a structural disadvantage.

III. Public Procurement and Automated Tendering

Automated procurement is particularly sensitive.

Suppose a government procurement platform automatically ranks suppliers according to an algorithm.

The algorithm may consider:

• price;
• previous government contracts;
• delivery history;
• financial information;
• ratings;
• compliance;
• technical compatibility.

The system may unintentionally create incumbency advantages.

Competition questions

Authorities should therefore ask:

1. Are selection criteria objectively justified?
2. Can new entrants compete?
3. Is historical government experience weighted excessively?
4. Is the algorithm auditable?
5. Are suppliers able to challenge automated decisions?
6. Does the system favour particular technologies?
7. Are public contracts being artificially concentrated?

IV. Automated Bid-Rigging and Algorithmic Collusion

Automation can facilitate collusion between suppliers.

Algorithms can:

• monitor competitors' prices;
• respond automatically to price changes;
• maintain predetermined margins;
• identify deviations from coordinated conduct;
• implement punishment strategies.

The traditional cartel model assumes communication between competitors. Automated systems complicate this assumption.

Important distinction

There are at least three forms:

A. Explicit algorithmic collusion

Competitors deliberately agree to use algorithms to implement a cartel.

B. Algorithm-facilitated collusion

Competitors communicate or coordinate manually but use algorithms to execute the arrangement.

C. Tacit algorithmic coordination

Independent algorithms learn that maintaining parallel prices is profitable without an explicit agreement.

The third category presents the greatest doctrinal difficulty.

V. Public Platforms as Gatekeepers

A government-operated platform may become an unavoidable gateway to public markets.

Examples include:

• government procurement portals;
• public transportation platforms;
• digital identity infrastructure;
• public payment systems;
• licensing platforms;
• public-sector cloud marketplaces.

If participation in the platform is effectively mandatory, the platform may possess substantial bottleneck power.

Competition governance must therefore ensure:

• non-discriminatory access;
• transparent technical standards;
• interoperability;
• reasonable access conditions;
• portability;
• procedural fairness.

VI. Essential Facilities and Automated Public Infrastructure

Some automated public systems may constitute essential infrastructure.

For example:

Government-controlled digital infrastructure → mandatory access → private firms depend on infrastructure → refusal/discrimination affects competition.

This raises issues similar to the essential facilities doctrine.

However, essential-facility analysis generally requires careful examination of:

• indispensability;
• availability of alternatives;
• market power;
• exclusionary effect;
• objective justification;
• proportionality.

A public system should not automatically be classified as an essential facility merely because it is government-operated.

VII. Public Undertakings and Competition Neutrality

A major problem occurs when the government participates in a market through a public enterprise.

Suppose a state-owned enterprise receives:

• privileged access to government data;
• subsidised infrastructure;
• preferential procurement;
• regulatory exemptions;
• exclusive access to an automated government platform.

Private competitors may then face a structurally unequal competitive environment.

The principle of competitive neutrality seeks to ensure that government ownership itself does not provide an unjustified competitive advantage.

VIII. Interoperability and Open Standards

Automated public systems often depend on technical standards.

A government may choose:

• one cloud architecture;
• one digital identity standard;
• one payment interface;
• one AI vendor;
• one cybersecurity protocol.

A technologically closed system can produce vendor lock-in.

Competition governance should therefore consider:

Open standards

Allowing competing providers to interoperate.

Data portability

Allowing users or competing suppliers to transfer relevant data.

API access

Permitting technically justified access to interfaces.

Multi-vendor procurement

Avoiding unnecessary dependence upon a single supplier.

IX. Algorithmic Transparency

Competition authorities increasingly face the question:

How can competition law investigate an algorithm whose decision-making process is opaque?

Relevant mechanisms include:

• algorithmic audits;
• source-code inspection where legally justified;
• independent technical experts;
• data-access orders;
• testing;
• documentation requirements;
• model-risk assessments;
• explanation requirements.

Transparency does not necessarily mean publishing source code publicly.

It may instead mean enabling regulators and affected parties to understand:

• inputs;
• decision criteria;
• weighting;
• outputs;
• error rates;
• discriminatory effects;
• mechanisms for human review.

X. Data as a Competitive Asset

In automated public systems, data can become an important competitive resource.

A public authority may have a unique dataset that private competitors cannot reproduce.

Competition issues arise when:

1. one undertaking receives exclusive access;
2. public data is licensed selectively;
3. data generated by private suppliers is subsequently used to favour another supplier;
4. interoperability is denied;
5. data portability is restricted.

Thus, data governance and competition governance increasingly overlap.

XI. Relevant Case Laws

The following cases provide important principles for analysing competition governance in automated or digitally mediated public systems.

1. Microsoft Corp. v Commission

Court: General Court of the European Union, 2007

Microsoft was found to have abused its dominant position through conduct involving interoperability information and tying.

Principle

Dominant firms controlling important technological infrastructure may have competition-law obligations concerning interoperability.

Relevance

The case is useful for automated public systems because a dominant digital infrastructure provider may control an interface or technical standard on which competitors depend.

It supports examination of:

• interoperability;
• access to technical information;
• technological foreclosure;
• tying;
• exclusion of competing systems.

2. Bronner v Mediaprint

Court: Court of Justice of the European Union, 1998

The case concerned access to a newspaper home-delivery network.

Principle

A refusal to provide access to infrastructure does not automatically constitute abuse of dominance. Strict conditions apply before compulsory access is required.

Relevance

Automated government platforms should not automatically be treated as essential facilities. Authorities must examine whether access is genuinely indispensable and whether viable alternatives exist.

3. IMS Health GmbH & Co. OHG v NDC Health

Court: Court of Justice of the European Union, 2004

The case concerned access to a data structure used in pharmaceutical-market information services.

Principle

Intellectual property and competition law can intersect where refusal of access to an indispensable resource substantially excludes competition.

Relevance

It is particularly relevant to automated public systems involving:

• proprietary databases;
• government datasets;
• interoperable information systems;
• digital infrastructure.

4. Slovak Telekom a.s. v Commission

Court: Court of Justice of the European Union, 2021

The case concerned access to telecommunications infrastructure and exclusionary conduct.

Principle

A dominant undertaking controlling infrastructure may face competition-law scrutiny where access conditions foreclose competitors.

Relevance

It provides an important analytical framework for public digital infrastructure where access to a network or platform determines competitors' ability to operate.

5. Google Shopping

Case: Google and Alphabet v Commission

EU Commission decision: 2017; General Court judgment: 2021

The case concerned preferential treatment of Google's comparison-shopping service in search results.

Principle

A dominant digital platform may engage in exclusionary conduct by using control over an important digital interface to favour its own service.

Relevance

The principle can be applied conceptually to automated public platforms where the platform operator:

• ranks suppliers;
• allocates opportunities;
• determines visibility;
• recommends providers;
• gives preferential treatment to affiliated services.

The key competition issue is self-preferencing through control over a digital gateway.

6. United States v. Google LLC

Court: U.S. District Court for the District of Columbia, 2024 merits decision

The case concerned Google's agreements and practices relating to distribution of general search.

Principle

The court examined how contractual arrangements involving distribution channels can reinforce market power and restrict competitive opportunities.

Relevance

The reasoning is useful for automated public systems where access to a government-controlled digital gateway or default position may determine which private providers obtain market exposure.

7. United States v. Microsoft Corp.

Court: U.S. Court of Appeals for the District of Columbia Circuit, 2001

The case involved Microsoft's conduct concerning the operating-system and browser markets.

Principle

Control over an important technological platform can enable a firm to disadvantage complementary or competing products.

Relevance

Automated public systems may similarly become platforms upon which multiple private providers depend. Competition analysis must therefore distinguish legitimate platform integration from exclusionary conduct.

8. MEO – Serviços de Comunicações e Multimédia

Court: Court of Justice of the European Union, 2018

The case concerned discriminatory pricing and the assessment of competitive disadvantage.

Principle

Discriminatory treatment by a dominant undertaking requires assessment of whether competition is actually or potentially distorted.

Relevance

Automated public allocation systems may use different prices, rankings, access conditions or eligibility thresholds. Competition law should examine whether differential treatment creates competitive disadvantage rather than assuming that every difference is unlawful.

XII. Automated Public Procurement: A Competition-Law Framework

A useful framework is:

Stage 1 — Identify the system

Determine whether the system is:

• a procurement platform;
• licensing platform;
• allocation mechanism;
• public marketplace;
• digital infrastructure;
• data platform.

Stage 2 — Identify the market

Determine:

• relevant product/service market;
• geographic market;
• upstream/downstream markets;
• adjacent digital markets.

Stage 3 — Identify the controller

Determine whether control lies with:

• government;
• public undertaking;
• private contractor;
• consortium;
• technology vendor.

Stage 4 — Examine market power

Consider:

• market share;
• network effects;
• switching costs;
• data advantages;
• interoperability;
• regulatory barriers;
• indispensability.

Stage 5 — Examine algorithmic conduct

Investigate:

• ranking;
• recommendation;
• exclusion;
• pricing;
• tender allocation;
• access decisions;
• default settings;
• interoperability restrictions.

Stage 6 — Examine competitive effects

Potential effects include:

• foreclosure;
• discriminatory access;
• raising rivals' costs;
• exclusion of new entrants;
• coordinated pricing;
• reduced innovation;
• concentration.

Stage 7 — Consider objective justification

Potential justifications may include:

• cybersecurity;
• public safety;
• privacy;
• fraud prevention;
• administrative efficiency;
• interoperability;
• national-security requirements.

Stage 8 — Select remedies

Possible remedies include:

• non-discriminatory access;
• interoperability;
• data portability;
• algorithmic audits;
• structural separation;
• procurement redesign;
• transparency obligations;
• behavioural commitments;
• divestiture in exceptional cases.

XIII. Automated Allocation and Public Resources

Automation increasingly determines access to scarce public resources.

Examples include:

• spectrum;
• airport slots;
• electricity-grid capacity;
• public housing;
• healthcare appointments;
• transport capacity;
• government contracts;
• environmental permits.

Competition law should ensure that allocation algorithms do not unnecessarily:

• exclude competitors;
• favour incumbents;
• create artificial scarcity;
• discriminate among suppliers;
• facilitate coordination.

XIV. Competition and Public-Private Partnerships

Automated public systems are frequently constructed through PPP arrangements.

A private technology provider may:

1. design the platform;
2. operate the infrastructure;
3. process government data;
4. determine algorithmic rules;
5. maintain the system;
6. supply complementary services.

This creates a potential concentration of infrastructure and informational power.

Competition governance should therefore consider:

• procurement neutrality;
• technology neutrality;
• subcontracting restrictions;
• data ownership;
• exit provisions;
• interoperability;
• switching rights;
• audit rights;
• anti-lock-in clauses.

XV. Automated Public Systems and Merger Control

Competition concerns can arise when firms supplying automated public infrastructure merge.

For example:

AI procurement platform + government-cloud provider + cybersecurity provider

may create vertical or conglomerate concerns.

Authorities may examine:

• foreclosure of rival suppliers;
• bundling;
• tying;
• access discrimination;
• data combination;
• interoperability restrictions;
• increased barriers to entry.

Traditional turnover thresholds may sometimes fail to capture strategically important digital transactions, making transaction-value or alternative jurisdictional thresholds relevant in some legal systems.

XVI. Algorithmic Collusion in Public Procurement

Public procurement is particularly vulnerable because repeated tenders produce large amounts of structured information.

Algorithms can observe:

• historical winning bids;
• losing bids;
• procurement schedules;
• supplier participation;
• regional allocation;
• contract duration.

A cartel could potentially use such information to coordinate:

• bid rotation;
• geographic allocation;
• market sharing;
• price levels.

Competition authorities should therefore monitor unusual patterns such as:

• repeated identical margins;
• predictable bid rotation;
• suspiciously stable winning sequences;
• identical pricing formulas;
• withdrawal patterns;
• algorithmically correlated bids.

Importantly, correlation alone does not establish an unlawful agreement. Investigators must distinguish legitimate algorithmic adaptation from concerted conduct.

XVII. Competition Governance and AI

Artificial intelligence adds several additional issues.

AI procurement

Governments purchasing AI systems should avoid unnecessarily exclusive technical requirements.

AI-generated procurement

AI tools should not inadvertently exclude smaller suppliers through biased historical data.

AI pricing

Suppliers using AI pricing systems can potentially facilitate coordinated outcomes.

AI vendor lock-in

Public authorities may become dependent upon a single foundation-model provider.

AI data advantages

A government dataset may provide significant competitive advantages to firms granted exclusive access.

XVIII. Indian Competition-Law Context

In India, automated public systems can be analysed principally through the Competition Act, 2002, alongside public-procurement, digital-governance, information-technology and sector-specific frameworks.

Relevant concepts include:

Section 3

Prohibits anti-competitive agreements.

This can encompass cartel arrangements and other agreements affecting competition.

Section 4

Addresses abuse of dominant position.

Potential automated-system issues include:

• discriminatory access;
• unfair conditions;
• denial of market access;
• leveraging;
• tying/bundling.

Sections 5 and 6

Govern combinations that may cause an appreciable adverse effect on competition.

Competition Commission of India

The CCI can therefore become relevant where an automated public system is operated by or through an undertaking engaging in economic activity.

XIX. Key Governance Principles

An effective competition-governance model should incorporate eight principles.

1. Competition neutrality

Public and private undertakings should compete under materially comparable conditions unless a justified public-policy distinction exists.

2. Non-discrimination

Automated systems should not arbitrarily favour particular suppliers.

3. Interoperability

Competing providers should be able to connect where technically and legally appropriate.

4. Data portability

Lock-in through control over data should be limited.

5. Algorithmic accountability

Important competitive decisions should be capable of regulatory examination.

6. Procedural fairness

Affected firms should have mechanisms to challenge significant automated decisions.

7. Proportionality

Restrictions justified by security, privacy or public-interest objectives should not unnecessarily suppress competition.

8. Regulatory coordination

Competition authorities should coordinate with:

• procurement authorities;
• data-protection regulators;
• sector regulators;
• digital regulators;
• cybersecurity authorities;
• public auditors.

XX. Competition Law vs Public-Interest Regulation

Automated public systems require careful separation of objectives.

The two objectives can overlap but are not identical.

A government may legitimately restrict competition for reasons such as national security or public safety. The competition question is whether the restriction is necessary, justified and appropriately designed, rather than whether competition is always maximised.

XXI. Emerging Doctrine: From Market Regulation to System Regulation

Traditional competition law primarily asks:

What did the undertaking do?

Automated public systems require an additional question:

How was the system designed, and what competitive incentives does that design create?

This moves competition governance toward system-level regulation.

The regulator may therefore need to investigate:

• architecture;
• interfaces;
• datasets;
• ranking mechanisms;
• default settings;
• APIs;
• access protocols;
• feedback loops;
• automated enforcement.

XXII. Practical Compliance Checklist

Before deploying an automated public system, authorities should ask:

Market structure

• Does the system create or reinforce a dominant position?
• Are alternatives available?

Procurement

• Can new entrants participate?
• Are technical specifications vendor-neutral?

Data

• Who controls the data?
• Is access discriminatory?

Algorithms

• What variables influence decisions?
• Can discriminatory outcomes be detected?

Interoperability

• Can rival providers connect?
• Are APIs reasonably accessible?

Lock-in

• Can the authority change suppliers?
• Can data be migrated?

Collusion

• Could the system facilitate coordination among suppliers?

Remedies

• Can decisions be challenged?
• Can the algorithm be audited?

XXIII. Overall Legal Significance

The principal competition-law challenge of automated public systems is that market power can be embedded in infrastructure rather than merely exercised through prices.

Control over:

• data,
• algorithms,
• interfaces,
• procurement gateways,
• technical standards,
• digital identity,
• cloud infrastructure,
• public platforms

may determine who can enter or survive in a market.

Consequently, future competition governance is likely to require a combination of:

competition law + procurement law + data governance + algorithmic accountability + interoperability regulation + sector regulation.

The most important lesson from the case law is that competition law does not automatically require public authorities or dominant infrastructure operators to provide unlimited access. Instead, the legal inquiry generally turns on market power, indispensability, discriminatory or exclusionary effects, competitive harm, objective justification, and proportionality.

Conclusion

Competition Law and Competition Governance in Automated Public Systems represents an evolution from conventional regulation of firms toward regulation of market-enabling technological architectures.

Automated public systems can deliver substantial benefits through efficiency, transparency, reduced administrative costs and faster allocation of resources. At the same time, they can create powerful competitive bottlenecks when a single authority or technology provider controls the data, infrastructure, algorithm and access mechanism.

The emerging legal framework therefore needs to ensure that automation does not become a mechanism for:

• exclusion;
• discriminatory access;
• algorithmic collusion;
• technological lock-in;
• self-preferencing;
• data foreclosure; or
• artificial barriers to entry.

The central principle can be expressed as:

Automation should make public markets more efficient without making them less contestable.