Competition Law and Algorithmic Employment Matching and Competition

1. Introduction

Algorithmic employment matching refers to the use of artificial intelligence, machine learning, automated ranking systems, and data analytics to connect workers with employers. These systems may identify suitable vacancies, rank applicants, recommend candidates to recruiters, determine which vacancies a worker sees, or decide which applicants proceed to the next stage.

Modern employment-matching systems can process CVs, skills, education, employment history, job descriptions, behavioural information, and other data to generate rankings or “fit” assessments. The EU’s AI framework specifically recognizes automated job-matching and ranking tools used in recruitment and selection as potentially high-risk AI systems.

From a competition-law perspective, the central issue is different from ordinary employment discrimination:

Does control over an employment-matching algorithm reduce competition among employers, restrict worker mobility, suppress compensation, exclude rival recruitment services, or allow competing employers to coordinate their conduct?

There is not yet a large body of decided antitrust cases specifically involving AI job-matching algorithms. Therefore, the legal framework is mainly constructed from established cases concerning labour-market monopsony, no-poach arrangements, wage information exchange, recruitment restrictions, platform power, and algorithmic coordination.

2. How Algorithmic Employment Matching Works

A simplified employment-matching ecosystem can be represented as:

Workers → Employment Platform → Matching Algorithm → Employers

The algorithm may consider:

• qualifications;
• experience;
• location;
• salary expectations;
• previous employment;
• employer preferences;
• historical hiring results;
• worker activity;
• application patterns;
• predicted performance;
• predicted likelihood of accepting an offer.

An algorithm can potentially improve competition because workers can discover more vacancies and employers can locate suitable candidates more efficiently.

Competition problems arise when the same technology becomes a gatekeeper controlling access between workers and employers.

For example:

Employer A + Employer B + Employer C → common matching platform → workers

If most employers rely on one platform, changes to its ranking algorithm could materially affect which employers and workers can find one another.

3. Relevant Competition-Law Principles

In the United States, the principal provisions are Sections 1 and 2 of the Sherman Act.

Section 1

Section 1 addresses agreements that unreasonably restrain trade.

In employment markets this can include agreements concerning:

• wages;
• employee recruitment;
• no-poach arrangements;
• allocation of workers;
• exchanges of sensitive compensation information.

An algorithm does not necessarily eliminate the competition issue simply because the restriction is implemented automatically.

Section 2

Section 2 addresses monopolization and attempted monopolization.

It can become relevant where a dominant employment platform allegedly uses exclusionary conduct to maintain market power.

For example, a platform controlling a substantial portion of online recruitment might theoretically disadvantage competing employment platforms by restricting access to essential recruitment data or employer integrations.

4. Labour Markets Are Competition Markets

Competition law does not protect only consumers purchasing products.

Workers sell their labour, while employers compete to acquire it.

Consequently:

Employers = buyers of labour

Workers = suppliers of labour

When employers possess substantial purchasing power over workers, economists commonly describe the situation as labour monopsony or oligopsony.

Algorithmic matching can affect this competitive process because algorithms increasingly determine which employers and workers encounter each other.

5. Case Law 1 — NCAA v. Alston

Case

National Collegiate Athletic Association v. Alston, 594 U.S. 69 (2021).

The case concerned NCAA restrictions affecting compensation and benefits available to student athletes.

The Supreme Court applied ordinary antitrust principles to the challenged restraints.

Principle

The decision is important because it confirms that restrictions affecting compensation in labour-related markets can receive meaningful antitrust scrutiny.

Algorithmic relevance

Imagine competing employers using the same employment-matching system.

The platform receives information concerning:

• salaries;
• bonuses;
• vacancies;
• hiring demand;
• acceptance rates.

Suppose its algorithm uses this collective information to recommend compensation levels.

If the arrangement reduced independent competition among employers for workers, ordinary antitrust principles concerning labour-market competition could become relevant.

Alston therefore provides an important foundation:

technology does not place labour-market restraints outside competition law.

6. Case Law 2 — Todd v. Exxon Corp.

Case

Todd v. Exxon Corp., 275 F.3d 191 (2d Cir. 2001).

Employees challenged an alleged exchange of compensation information among major employers.

The Second Circuit treated the exchange of competitively sensitive salary information as capable of raising antitrust concerns and emphasized analysis of market structure and competitive effects.

Algorithmic significance

This precedent is particularly relevant to shared employment algorithms.

Suppose several competing employers continuously submit:

• salaries;
• proposed salary increases;
• vacancies;
• recruitment activity;
• employee turnover;
• bonuses.

A common AI platform processes the information and produces recommendations for each employer.

The question becomes whether the platform has effectively created a centralized information-exchange mechanism.

The relevant factors could include:

1. how concentrated the labour market is;
2. whether the information is current or historical;
3. whether individual employers can be identified;
4. whether recommendations influence actual salaries;
5. whether employers continue making genuinely independent decisions.

Todd therefore provides an important framework for examining algorithmically mediated exchanges of labour-market information.

7. Case Law 3 — United States v. eBay, Inc.

Case

United States v. eBay, Inc., 968 F. Supp. 2d 1030 (N.D. Cal. 2013).

The U.S. government alleged that eBay and Intuit entered into an agreement restricting recruitment and hiring between the companies.

The court allowed the government's Sherman Act claim to proceed past the pleading stage.

Principle

Competing employers can compete against each other for employees.

Restrictions preventing them from recruiting each other's workers can therefore restrict labour-market competition.

Algorithmic employment matching

Consider two employers using an automated recruitment platform.

Instead of managers directly agreeing not to recruit each other's employees, the platform is configured to:

detect current employer → suppress candidate → prevent recommendation to competing employer.

The technological implementation would not automatically resolve the antitrust issue.

Competition authorities would examine whether an underlying agreement existed and whether it restricted competition for workers.

8. Case Law 4 — Deslandes v. McDonald's USA, LLC

Case

Deslandes v. McDonald's USA, LLC, 81 F.4th 699 (7th Cir. 2023).

The litigation concerned restrictions affecting recruitment among McDonald's franchise restaurants.

The Seventh Circuit rejected treating the restraint as automatically protected merely because it appeared within a franchise arrangement and required closer consideration of whether the restriction was properly ancillary to the legitimate collaboration.

Algorithmic relevance

A franchise recruitment platform could automatically prevent:

Franchise A → recruiting employee from Franchise B.

Such a rule might be embedded directly within software.

Competition analysis would nevertheless ask whether the recruitment restriction is a legitimate ancillary restraint or instead constitutes an unnecessary restriction on competition for workers.

Therefore:

automating a no-poach restriction does not determine its legality.

9. Case Law 5 — Aya Healthcare Services, Inc. v. AMN Healthcare, Inc.

Case

Aya Healthcare Services, Inc. v. AMN Healthcare, Inc., 9 F.4th 1102 (9th Cir. 2021).

The dispute involved competing healthcare staffing businesses and a contractual non-solicitation restriction.

The Ninth Circuit treated the restraint as ancillary to a legitimate collaborative relationship rather than automatically condemning it as a naked market-allocation arrangement.

Importance

The case demonstrates that not every employment-related restriction violates competition law.

Courts distinguish between:

Naked restraint: a restriction primarily eliminating competition.

and

Ancillary restraint: a reasonably necessary restriction connected with a legitimate business collaboration.

Algorithmic example

Two staffing businesses could operate a legitimate joint recruitment system.

Certain limited restrictions might be necessary to operate that collaboration.

But if their shared algorithm broadly prevented either business from recruiting the other's employees across unrelated activities, competition concerns would become considerably different.

10. Case Law 6 — Bogan v. Hodgkins

Case

Bogan v. Hodgkins, 166 F.3d 509 (2d Cir. 1999).

The dispute involved restrictions affecting movement of insurance agents.

The case illustrates the importance of determining whether the challenged arrangement actually constitutes a meaningful allocation or restriction of competition within a relevant labour market.

Algorithmic significance

Suppose an AI matching system refuses to recommend workers currently employed by certain competing firms.

The existence of this technical restriction alone does not answer the antitrust question.

Analysis would examine:

• which workers are affected;
• which employers compete for those workers;
• alternative employment opportunities;
• geographic scope;
• occupational specialization;
• duration of the restriction;
• actual effects on recruitment and compensation.

This illustrates why labour-market definition is particularly important in algorithmic employment cases.

11. Case Law 7 — United States v. RealPage, Inc.

The government's litigation against RealPage concerns rental housing rather than employment matching, but it provides a useful modern analogy for algorithmic competition questions.

The government has alleged that competing landlords supplied non-public information to common software used for pricing recommendations, raising Sherman Act Sections 1 and 2 issues.

For employment markets, the analogous structure would be:

Employers → confidential wage/hiring information → common algorithm → compensation or recruitment recommendations.

The important analytical question would be whether employers remained independent competitors or whether the common system facilitated coordination.

The RealPage litigation therefore illustrates how established antitrust principles may be applied where algorithms sit between competing businesses.

12. Case Law 8 — Mobley v. Workday, Inc.

Case

Mobley v. Workday, Inc., No. 23-cv-00770-RFL (N.D. Cal.).

This is particularly important for understanding automated employment screening.

Applicants alleged that Workday's algorithm-based applicant-screening tools discriminated against applicants based on protected characteristics. In 2024, the district court allowed significant parts of the case to proceed, including under an agency theory.

The litigation continued into 2026, with allegations describing automated systems used for screening, scoring and ranking applicants.

Important distinction

Mobley is principally an employment-discrimination case, not an antitrust judgment establishing that algorithmic matching violates competition law.

Nevertheless, it demonstrates something important for future competition cases: centralized employment technology can have substantial influence over access to employment opportunities.

Competition-law analysis would ask different questions, including whether such a platform possesses market power or uses its position to exclude competing recruitment services or suppress competition among employers.

13. Algorithmic Labour-Market Allocation

One particularly serious hypothetical competition problem would occur if competing employers used matching algorithms to divide workers.

For example:

Employer A receives engineers

Employer B receives designers

Employer C receives analysts

If this division resulted from an agreement among competitors rather than genuine competitive matching, it could resemble traditional market allocation.

The fact that software executes the allocation would not necessarily change the underlying economic character of the arrangement.

14. Algorithmic No-Poach Systems

Algorithms could also implement recruitment restrictions automatically.

For example:

Candidate employed by Company A

↓

Matching algorithm detects employer

↓

Candidate excluded from Company B's recommendations

If Company A and Company B independently make legitimate hiring decisions, there may be no horizontal agreement.

But if competing companies agree that their common system should prevent recruitment from one another, Section 1 concerns can arise.

The distinction between independent conduct and coordinated conduct is therefore crucial.

15. Wage Coordination

Employment-matching platforms can possess unusually valuable information about labour demand.

They may know:

• current salaries;
• salary expectations;
• rejected offers;
• accepted offers;
• worker shortages;
• application volumes;
• employer demand;
• employee turnover.

These datasets can improve matching efficiency.

But they could also create competition concerns where competitively sensitive information from rival employers is pooled and used to influence their future compensation decisions.

This is where Todd v. Exxon becomes particularly relevant.

16. Platform Gatekeeper Power

A large employment platform can connect two groups:

workers ↔ employers

This creates network effects.

More workers attract employers.

More employers attract workers.

Consequently, successful platforms can become difficult for new recruitment platforms to challenge.

Competition authorities may therefore examine:

• network effects;
• switching costs;
• data advantages;
• interoperability;
• exclusive contracts;
• employer multihoming;
• worker multihoming;
• API restrictions.

These factors can affect whether a platform possesses substantial market power.

17. Self-Preferencing

Suppose an employment platform also operates its own staffing business.

Its algorithm might theoretically:

rank its staffing service first

while

downgrading competing staffing agencies.

Competition analysis would require evidence concerning market power, exclusionary effects, justification, and the governing jurisdiction's legal standard.

Self-preferencing is therefore not automatically an antitrust violation, but under some circumstances it can form part of an exclusionary-conduct theory.

18. Data as a Competitive Advantage

Employment platforms can accumulate enormous datasets concerning:

worker → skills → applications → interviews → offers → salaries → hiring outcomes.

Historical matching data may improve future algorithmic recommendations.

This can create a feedback loop:

more users → more data → better matching → more employers → more workers → more data.

Such effects can produce legitimate efficiencies.

However, competition authorities may investigate whether a dominant firm uses contractual, technical, or exclusionary practices to prevent competitors from obtaining inputs needed to compete effectively.

19. Algorithmic Discrimination vs Competition Harm

These concepts should be kept separate.

Employment discrimination

The question is whether applicants are treated unlawfully because of characteristics protected by employment law.

Competition law

The question is whether competitive conditions in a relevant market have been harmed.

Therefore, an algorithm could potentially be discriminatory without creating an antitrust violation.

Likewise, an algorithm could potentially restrict competition without discriminating against a legally protected class.

Mobley v. Workday illustrates the first category: the litigation focuses principally on alleged discrimination in automated applicant screening.

20. Relevant Market Definition

An antitrust investigation would ordinarily need to determine the relevant market.

Possible labour markets could include:

Software engineers in a particular geographic area

rather than:

all workers in the United States.

Relevant factors include:

• worker skills;
• qualifications;
• occupation;
• geography;
• ability to switch occupations;
• employer alternatives;
• remote-working possibilities.

Alternatively, authorities could investigate a market for online employment-matching services rather than the underlying labour market.

The appropriate market depends on the alleged competitive harm.

21. Monopsony Power

Algorithmic employment matching can also interact with monopsony.

Suppose only a small number of employers hire workers with highly specialized skills.

If a dominant matching platform enables those employers to coordinate recruitment or compensation, workers could have fewer realistic alternatives.

Potential effects could include:

lower wages + reduced mobility + fewer job opportunities + weaker employer competition.

NCAA v. Alston is important because modern antitrust doctrine clearly recognizes competitive harm occurring on the labour side of markets.

22. Potential Procompetitive Benefits

Algorithmic employment matching can also substantially improve competition.

Potential efficiencies include:

• faster recruitment;
• lower search costs;
• better worker-job matching;
• discovery of previously unknown vacancies;
• access to geographically distant employers;
• improved skill matching;
• lower recruitment costs for small businesses;
• faster identification of labour shortages.

Competition law therefore should not assume that employment algorithms are inherently harmful.

The central issue is how the technology affects the competitive process.

23. Main Competition Risks

The major competition-law risks can be summarized as:

24. Case-Law Summary

Conclusion

Algorithmic employment matching is not inherently anti-competitive. Properly designed systems can increase labour-market competition by reducing search costs and connecting workers with more employers.

Competition-law concerns become more significant where algorithms are used to coordinate wages, implement no-poach arrangements, allocate workers, exchange competitively sensitive employment information, exclude rival recruitment platforms, or reinforce substantial labour-market or platform power.

Existing cases such as NCAA v. Alston, Todd v. Exxon, United States v. eBay, Deslandes v. McDonald's, Aya Healthcare v. AMN, and Bogan v. Hodgkins supply much of the established legal framework. RealPage provides a modern algorithmic-coordination analogy, while Mobley v. Workday demonstrates the growing practical significance of centralized automated employment screening.

The fundamental competition-law principle remains technologically neutral: when employers are supposed to compete independently for workers, using an algorithm does not by itself remove ordinary antitrust scrutiny.