Competition Law and Process Automation Ecosystems and Competition

1. Introduction

Process automation ecosystems are digital environments in which software, algorithms, artificial intelligence, APIs, workflow engines, robotic process automation (RPA), cloud services, enterprise applications, and data systems automate business processes.

Examples include automated:

procurement;

invoicing;

payments;

logistics;

customer service;

inventory management;

compliance;

recruitment;

insurance processing;

manufacturing;

accounting; and

supply-chain management.

Automation can significantly increase productivity and reduce transaction costs. However, where a small number of technology providers control the underlying automation infrastructure, competition concerns can emerge.

The central competition-law question is:

When does process automation improve competitive efficiency, and when can control over automated processes become a mechanism for coordination, exclusion, foreclosure, or market concentration?

2. Structure of a Process Automation Ecosystem

A typical ecosystem may contain several layers:

Infrastructure layer

Cloud computing, databases, computing infrastructure.

↓

Automation layer

RPA, workflow engines, AI models, orchestration systems.

↓

Application layer

ERP, CRM, procurement, accounting and logistics applications.

↓

API/interface layer

Connections between applications and external providers.

↓

Business users

Manufacturers, banks, retailers, hospitals, governments and other enterprises.

↓

Consumers and downstream markets

The competitive effects can arise at any of these levels.

3. Why Competition Law Matters

Automation changes competition because it can:

lower operating costs;

reduce labour requirements;

improve accuracy;

increase speed;

standardise business practices;

create economies of scale;

increase switching costs;

concentrate data;

facilitate interoperability; and

make businesses dependent on particular software providers.

At the same time, automation can potentially facilitate:

algorithmic collusion;

exclusionary conduct;

tying;

interoperability restrictions;

discriminatory access;

self-preferencing;

technological foreclosure; and

acquisition of emerging competitors.

4. Relevant Competition-Law Framework

Process automation ecosystems can implicate several traditional competition-law doctrines.

Article 101 TFEU / Section 3 Competition Act

These provisions address anti-competitive agreements and concerted practices.

Relevant conduct may include:

price fixing;

market allocation;

information exchange;

bid rigging;

restrictions on technological access; and

coordination through common automation systems.

Article 102 TFEU / Section 4 Competition Act

These provisions concern abuse of dominance.

Possible theories include:

refusal to supply;

discriminatory access;

tying;

bundling;

self-preferencing;

exclusionary pricing;

interoperability restrictions; and

leveraging dominance from one market into another.

5. Automation and Algorithmic Collusion

One of the most important issues is whether automation makes coordination easier.

Suppose competing manufacturers use the same automated pricing system.

The system continuously observes:

market prices;

demand;

competitor behaviour;

inventory; and

price changes.

It automatically adjusts prices.

The result could be highly stable parallel pricing.

However:

Parallel pricing by algorithms does not automatically establish an unlawful cartel.

Competition authorities would need to establish the legally relevant elements of an infringement under the applicable legal system.

The risk nevertheless increases where competitors deliberately configure systems to coordinate prices or exchange competitively sensitive information.

6. Common Automation Provider as a Coordination Hub

A particularly important scenario involves a common automation provider.

For example:

Retailer A → Automation Platform

Retailer B → Automation Platform

Retailer C → Automation Platform

The platform collects pricing and inventory data and generates recommendations for each retailer.

If competitively sensitive information is transmitted between competitors, the automation platform may potentially become a hub through which coordination occurs.

This raises questions concerning:

information flows;

platform design;

knowledge of the participants;

purpose;

foreseeability;

algorithmic instructions; and

resulting market behaviour.

7. Information Exchange

Automation systems frequently process enormous amounts of commercial data.

Potentially sensitive information includes:

prices;

costs;

production levels;

inventory;

future strategy;

capacity;

customer information;

procurement plans; and

anticipated demand.

A platform that aggregates such information must therefore distinguish between information necessary to provide an efficient service and information whose disclosure could reduce competitive uncertainty.

8. Dominance in Automation Infrastructure

A process-automation provider can become dominant because customers invest heavily in:

software integration;

employee training;

data migration;

custom workflows;

APIs;

enterprise contracts; and

system compatibility.

This creates switching costs.

Once an enterprise automates thousands of processes around a particular platform, changing providers may require substantial expenditure.

The result can be technological lock-in.

A dominant provider could potentially exploit this dependence through:

excessive fees;

restrictive licensing;

discriminatory interoperability;

tying;

exclusive arrangements; or

refusal to provide necessary interfaces.

9. Network Effects

Automation ecosystems can generate network effects.

More users produce:

more data → better automation → greater attractiveness → more users → more data.

This feedback loop can reinforce concentration.

For example, an automation platform serving millions of businesses may possess data concerning:

transaction patterns;

workflow efficiency;

supplier performance;

demand;

operational costs; and

business processes.

A new entrant may struggle to reproduce the same data advantage.

10. Data as an Entry Barrier

Data can become a competitive asset when it improves automation models.

A dominant automation provider might possess:

historical workflow data;

performance data;

industry benchmarks;

customer behaviour;

transaction data; and

machine-learning training data.

Competitors without equivalent datasets may find entry more difficult.

Nevertheless, data possession alone does not establish an antitrust violation.

The analysis must consider:

substitutability;

replicability;

market power;

access conditions;

competitive effects; and

legitimate business justifications.

11. Interoperability and APIs

Automation ecosystems depend heavily upon APIs.

An API allows one system to communicate with another.

For example:

ERP → API → Automation Platform → Payment System

A dominant platform could potentially restrict APIs to prevent competitors from integrating.

This creates possible competition concerns involving:

refusal to supply;

interoperability restrictions;

discrimination;

technological foreclosure; and

leveraging.

Where a rival cannot effectively compete without interoperability, access conditions may become an important part of the competition analysis.

12. Tying and Bundling

A dominant automation provider may attempt to combine several products:

Automation software + cloud services + analytics + security + payment services.

Bundling is not inherently unlawful.

The competition issue arises where a dominant firm uses power in one market to disadvantage competitors in another.

For example:

Dominant workflow software

↓

Mandatory use of

proprietary cloud infrastructure

↓

Competitors cannot provide competing cloud services.

Such conduct could raise tying or leveraging concerns depending on the applicable legal framework and competitive effects.

13. Self-Preferencing

Suppose an automation platform operates a marketplace for third-party business applications.

The platform also owns competing applications.

Its automated recommendation engine could systematically favour its own applications.

For example:

Platform's own accounting software → Priority recommendation

Independent accounting software → Lower ranking

The issue resembles the concerns examined in the Google Shopping litigation.

Algorithmic preference can therefore become an important form of competitive conduct.

14. Refusal to Interoperate

A dominant automation ecosystem may refuse to allow competitors to connect to:

APIs;

databases;

authentication systems;

workflow engines;

cloud environments; or

payment infrastructure.

The legal analysis may involve the principles associated with essential facilities and refusal-to-supply cases.

However, not every refusal to provide access is unlawful.

Courts generally examine factors such as:

indispensability;

elimination of competition;

objective justification;

feasibility of access; and

competitive effects.

15. Case Law 1 — United Brands v Commission

Case 27/76, Court of Justice of the European Union

United Brands is a foundational dominance case.

The Court considered the behaviour of a dominant undertaking and the conditions under which commercial conduct can constitute abuse.

Relevance to automation

A dominant process-automation provider may possess substantial power over dependent business users.

The case provides foundational principles for analysing:

dominance;

unfair conditions;

discriminatory treatment;

market power; and

exclusionary conduct.

16. Case Law 2 — Bronner v Mediaprint

Case C-7/97, CJEU

The Court examined refusal of access to an infrastructure controlled by a dominant undertaking.

Relevance

Process automation increasingly depends upon digital infrastructure.

If a dominant automation provider controls an indispensable technical interface, questions may arise regarding whether refusing access can exclude competitors.

The Bronner framework is therefore relevant to digital interoperability disputes.

17. Case Law 3 — Microsoft Corp. v Commission

Case T-201/04, General Court

Microsoft's conduct concerning interoperability information and the tying of Windows with Windows Media Player became an important EU competition-law precedent.

Relevance to process automation

The case is especially relevant because modern automation ecosystems frequently involve:

interoperability;

software integration;

technical interfaces;

tying;

bundling; and

platform leverage.

A dominant automation provider that restricts interoperability or conditions access to one product on adoption of another could raise comparable legal issues.

18. Case Law 4 — Google Shopping

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

The case is important for understanding self-preferencing through algorithmic systems.

Relevance

Automation platforms increasingly determine:

supplier ranking;

application recommendations;

workflow recommendations;

software selection; and

resource allocation.

If a dominant platform systematically favours its own downstream services, algorithmic self-preferencing can become an important competition-law issue.

19. Case Law 5 — Intel Corp. v European Commission

Case C-413/14 P, CJEU

The Intel litigation concerned alleged exclusionary conduct by a dominant undertaking.

The CJEU emphasised the importance of examining relevant economic and competitive effects in appropriate circumstances.

Relevance

Automation providers may use:

loyalty incentives;

exclusive contracts;

rebates;

bundled services;

preferential access; or

technical restrictions.

The Intel jurisprudence is useful for analysing whether such practices actually have the capacity to foreclose competitors.

20. Case Law 6 — Eturas UAB and Others v Lietuvos Respublikos konkurencijos taryba

Case C-74/14, CJEU

This case concerned an electronic booking platform through which information about discount restrictions was communicated to participating businesses.

Importance

It is particularly relevant to competition in automated digital environments because the conduct occurred through an electronic platform.

Relevance to automation

The case illustrates how an electronic system can become relevant to the assessment of concerted practices.

A common automation platform that communicates competitively sensitive information between competitors could therefore create similar competition-law questions.

21. Case Law 7 — T-Mobile Netherlands v NMa

Case C-8/08, CJEU

The case concerned information exchange between competitors.

Relevance

Automation systems can make information exchange continuous and instantaneous.

Instead of competitors physically meeting, information may be exchanged through:

APIs;

cloud systems;

shared databases;

algorithmic platforms; or

automated reporting systems.

The underlying competition-law concern remains the reduction of uncertainty between competitors.

22. Case Law 8 — Aalborg Portland and Others v Commission

Joined Cases C-204/00 P and others, CJEU

The case concerns proof of participation in cartel conduct and the assessment of coordinated behaviour.

Relevance

Automated ecosystems create extensive electronic evidence.

Competition authorities may examine:

software configurations;

communications;

transaction records;

platform logs;

algorithmic instructions; and

user activity.

Aalborg Portland is relevant to the broader evidentiary principles surrounding participation in coordinated conduct.

23. Automation and Procurement

Process automation is particularly significant in procurement.

An automated procurement platform may:

identify suppliers;

predict demand;

evaluate bids;

determine purchase quantities;

recommend prices; and

automatically award contracts.

Potential competition risks include:

bid rigging;

supplier exclusion;

discriminatory ranking;

coordinated pricing;

preferential treatment;

information exchange; and

foreclosure of smaller suppliers.

Automation therefore can simultaneously improve procurement efficiency and create new channels for anticompetitive conduct.

24. Automation and Labour Markets

Process automation also affects labour markets.

Where a small number of technology providers control workforce-management systems, concerns may arise regarding:

wage-setting algorithms;

worker allocation;

recruitment platforms;

automated scheduling;

worker classification; and

sharing of compensation information.

If competing employers use a common system that facilitates coordination concerning wages or employment conditions, competition authorities may examine whether the system contributes to unlawful coordination.

This area is increasingly important because competition law can apply to purchasing power and employer-side conduct, not merely product markets.

25. Automation and Consumer Markets

Automation may influence consumers through:

personalised prices;

automated recommendations;

dynamic pricing;

personalised promotions;

automated customer segmentation; and

subscription management.

A dominant platform could potentially use automated systems to discriminate among customers or disadvantage competitors.

The legality would depend on the applicable competition-law framework and the existence of dominance, agreement, or other legally relevant conduct.

26. Automated Pricing

Automated pricing systems can create two distinct scenarios.

Scenario A — Independent optimisation

Each company independently uses its own algorithm to respond to market conditions.

This may simply constitute competition using technology.

Scenario B — Coordinated optimisation

Competitors deliberately use systems designed to align prices or exchange competitively sensitive information.

This can create substantially greater competition-law risk.

The distinction between these scenarios is therefore critical.

27. Tacit Coordination

The most difficult problem may arise where algorithms independently learn to avoid aggressive competition.

For example:

Firm A's algorithm raises price

↓

Firm B's algorithm detects the change

↓

Firm B automatically raises price

↓

Firm A detects B's response

↓

Repeated algorithmic interaction produces stable prices

There may be no explicit communication.

Competition law must therefore distinguish:

lawful conscious adaptation to market conditions;

tacit coordination;

concerted practice; and

explicit cartel conduct.

The legal treatment differs considerably across jurisdictions.

28. Automation and Market Concentration

Automation can create economies of scale.

A platform that serves millions of businesses may spread software-development costs over a very large customer base.

This can produce lower average costs.

However, scale can also create concentration.

Large providers may gain advantages through:

data;

integration;

network effects;

brand;

cloud infrastructure;

customer switching costs; and

accumulated machine-learning capabilities.

Competition authorities therefore need to determine whether scale reflects legitimate efficiency or whether it is reinforced by exclusionary practices.

29. Killer Acquisitions in Automation

Large technology companies may acquire small automation startups.

An acquisition may concern a company that:

has limited current revenue;

possesses valuable technology;

has important data;

has rapidly growing users; or

could become a future competitor.

Traditional turnover thresholds may fail to capture the competitive significance of such acquisitions in some circumstances.

Consequently, merger-control authorities may examine:

innovation competition;

potential competition;

data assets;

technology;

ecosystem effects; and

future competitive constraints.

30. Interoperability as a Competition Remedy

Where automation concentration creates competitive concerns, interoperability can become an important remedy.

Possible measures include:

API access;

data portability;

technical standards;

non-discriminatory integration;

interoperability requirements; and

restrictions on discriminatory technical barriers.

The objective is to allow users to connect competing services without unnecessarily dismantling the underlying platform.

31. Indian Competition-Law Perspective

Under India's Competition Act, 2002, process-automation ecosystems may raise issues principally under Sections 3 and 4.

Section 3

Potential concerns include:

price-fixing through automated systems;

bid rigging;

information exchange;

market allocation;

coordination through common software; and

restrictive technological agreements.

Section 4

Where an automation provider holds a dominant position, possible concerns include:

discriminatory access;

denial of market access;

tying;

bundling;

leveraging;

exclusionary contracts;

discriminatory interoperability; and

other forms of abusive conduct.

Merger control

Acquisitions involving major automation platforms may also raise questions under India's combination-control framework where the statutory thresholds and substantive requirements are satisfied.

32. Role of Competition Authorities

Competition authorities can themselves use process automation.

Automated systems can identify:

suspicious tender patterns;

unusual price movements;

coordinated bidding;

supplier rotation;

exclusionary patterns;

discriminatory access;

market concentration; and

abnormal algorithmic behaviour.

Thus, automation creates a dual phenomenon:

The same technological architecture can facilitate competition enforcement and potentially facilitate competition violations.

33. Compliance Framework for Automation Platforms

Businesses operating process-automation ecosystems can reduce competition risks through:

1. Information controls

Restrict access to competitively sensitive information.

2. Algorithmic governance

Document how pricing, ranking, and allocation algorithms operate.

3. Competition-law audits

Periodically assess automated processes.

4. Interoperability policies

Avoid unjustified restrictions on legitimate integrations.

5. Non-discrimination

Apply platform rules consistently to competing users.

6. Human oversight

Maintain appropriate review of high-impact automated decisions.

7. Data separation

Prevent inappropriate sharing of competitors' confidential information.

8. Acquisition review

Assess competition implications before acquiring emerging automation competitors.

34. Key Competition Questions

When analysing an automation ecosystem, the following questions should be asked:

Who controls the automation infrastructure?

Does the provider possess substantial market power?

Are customers locked into the platform?

Can competing providers interoperate?

Does the platform favour its own products?

What competitively sensitive information does it possess?

Can competitors access each other's information?

Can algorithms coordinate market behaviour?

Does the system create barriers to entry?

Are there legitimate efficiency justifications for the challenged conduct?

35. Competition Risks and Benefits

36. Conclusion

Process automation ecosystems create a new layer of competition-law complexity because technology increasingly determines how businesses interact with markets.

The principal competition concerns are:

algorithmic coordination;

information exchange;

platform dominance;

interoperability restrictions;

self-preferencing;

tying and bundling;

data-driven entry barriers;

customer lock-in;

exclusionary automation; and

concentration through acquisitions.

The cases of United Brands, Bronner, Microsoft, Google Shopping, Intel, Eturas, T-Mobile Netherlands, and Aalborg Portland demonstrate that traditional competition principles remain highly relevant even when commercial conduct is implemented through sophisticated digital systems.

The central issue is therefore not whether automation itself is pro-competitive or anti-competitive. Automation can simultaneously produce substantial efficiencies and amplify market power. Competition-law analysis must examine the structure of the ecosystem, the conduct of the participants, information flows, interoperability, market power, and actual or potential effects on competitive conditions.

In the Indian context, Sections 3 and 4 of the Competition Act, 2002 provide the principal framework for assessing agreements, coordinated conduct, and abuse of dominance arising from process-automation ecosystems. The growing importance of AI, APIs, cloud infrastructure, automated decision-making, and enterprise software means that competition law will increasingly have to address not merely who controls the market, but who controls the automated processes through which the market operates.