AI-Generated Media and Competition Displacement of Human Creators

1. Introduction

Artificial intelligence has transformed the production of media by enabling automated generation of text, music, photographs, illustrations, video, voice performances, advertisements, news content and other creative works. Generative AI systems can produce substantial quantities of media at relatively low marginal cost and at much greater speed than conventional human creators.

From a competition-law perspective, the central issue is not simply whether AI-generated content is “better” or “worse” than human-created content. The concern is whether the deployment of AI changes competitive conditions by displacing human creators, increasing concentration among firms controlling AI models or distribution platforms, reducing access to audiences, facilitating exclusionary conduct, or enabling large firms to reproduce creative outputs without equivalent costs.

Competition law therefore intersects with copyright, labour markets, platform regulation, data governance and consumer protection.

2. Meaning of AI-Generated Media

AI-generated media refers to content substantially produced or transformed through machine-learning or generative-AI systems.

It may include:

1. AI-generated text – articles, scripts, books and advertising copy.
2. AI-generated images – illustrations, photographs and graphic designs.
3. AI-generated music – compositions, background music and synthetic performances.
4. AI-generated video – advertisements, films, animation and synthetic actors.
5. AI-generated voices – narration, dubbing and voice cloning.
6. AI-generated journalism – automated news reports and summaries.
7. AI-generated advertising – automatically created campaigns and promotional material.
8. Synthetic influencers and virtual creators.
9. AI-assisted professional creative work – editing, post-production, translation and design.

The competitive significance increases where AI becomes a substitute for previously human-performed creative services.

3. Competition Displacement of Human Creators

AI can affect competition through several interconnected mechanisms.

A. Direct substitution

A company may replace human designers, illustrators, copywriters, translators, musicians or editors with AI systems.

The immediate competitive effect can be a reduction in demand for particular categories of human creative services.

B. Cost asymmetry

A human creator may require:

• wages or fees;
• working hours;
• studio equipment;
• software;
• production expenses;
• contractual payments; and
• repeated revisions.

An AI system may generate thousands of outputs at comparatively low incremental cost.

This creates a significant cost asymmetry between automated and human production.

C. Speed advantage

AI can generate content almost instantaneously.

A platform capable of producing 10,000 advertising images through automation may compete against individual creators who can produce only a small number of comparable works within the same period.

D. Platform dependence

Human creators frequently depend upon:

• social-media platforms;
• search engines;
• streaming services;
• online marketplaces;
• advertising exchanges;
• app stores; and
• creator platforms.

If the same platform controls AI-generation tools and distribution, it may possess an additional competitive advantage.

4. Relevant Competition-Law Issues

A. Market Definition

A competition authority may need to determine whether AI-generated and human-generated media belong to:

1. the same relevant product market;
2. partially substitutable markets; or
3. separate markets.

For example, AI-generated stock images may compete directly with human-created stock photography, while AI-generated feature films may initially be complementary rather than complete substitutes for human-produced films.

The relevant inquiry may consider:

• price;
• quality;
• consumer preferences;
• speed;
• originality;
• authenticity;
• contractual requirements;
• copyright status; and
• switching possibilities.

5. Predatory or Below-Cost AI Content

A dominant technology or media company could theoretically use AI-generated content to offer services below the sustainable cost of human competitors.

For example, a platform might provide:

AI-generated illustrations for zero or near-zero monetary price while simultaneously controlling the distribution channel through which professional illustrators reach customers.

If this conduct forms part of a broader exclusionary strategy, authorities could examine it under abuse-of-dominance principles.

The existence of inexpensive AI output, however, is not by itself an antitrust violation. Competition law generally protects the competitive process rather than guaranteeing that existing production methods remain economically viable.

6. AI and Entry Barriers

AI can simultaneously reduce and increase barriers to entry.

Reducing barriers

A small creator can use AI for:

• editing;
• translation;
• animation;
• voice generation;
• graphic design;
• marketing; and
• content production.

This may allow small businesses to compete with larger firms.

Increasing barriers

At the infrastructure level, however, sophisticated generative AI may require:

• enormous datasets;
• GPUs;
• cloud infrastructure;
• computing capacity;
• foundation models;
• specialised engineers;
• distribution platforms; and
• substantial capital.

Consequently, AI may reduce barriers at the content-production level while increasing concentration at the AI-infrastructure level.

7. Control of Training Data

Training data can become a critical competitive input.

Large AI developers may possess access to enormous repositories of:

• books;
• photographs;
• music;
• videos;
• news articles;
• online posts;
• databases; and
• professional creative works.

If access to high-quality training data becomes indispensable, control over such datasets may produce a competitive advantage.

The competition concern becomes stronger where a dominant undertaking:

1. controls an important dataset;
2. prevents competitors from obtaining equivalent data;
3. uses the data to improve its own AI system; and
4. thereby reinforces its market position.

8. Data Scraping and Competitive Harm

The collection of human-created works for AI training raises a distinctive competition issue.

Suppose a dominant platform:

• hosts millions of creators;
• collects their content;
• trains its proprietary AI system on that content;
• develops an AI substitute for those creators; and
• subsequently competes against the same creators.

This may create a potential vertical conflict of interest.

The platform could effectively transform its users' creative output into an input for a competing automated service.

Whether this constitutes an antitrust infringement depends on market power, conduct, contractual arrangements, applicable intellectual-property rights and competitive effects.

9. Self-Preferencing

A platform that both distributes human-created media and generates AI content may have incentives to prefer its own AI-produced material.

Potential practices include:

• higher search rankings for AI-generated content;
• preferential recommendation;
• cheaper advertising;
• greater visibility;
• preferential monetisation;
• exclusion from competing AI systems; or
• restricting competing creator tools.

This resembles broader concerns concerning platform self-preferencing.

10. Bundling and Tying

A dominant platform could tie:

AI content-generation tools + distribution + advertising + analytics.

For example, a platform might require creators to use its proprietary AI-generation system to obtain access to premium distribution.

Competition authorities may examine whether such conduct forecloses competing creative or AI-service providers.

11. Exclusive Dealing

AI companies or media platforms may enter exclusive arrangements with:

• actors;
• musicians;
• publishers;
• studios;
• influencers;
• news organisations; or
• professional creator associations.

Exclusive access to particularly valuable datasets or talent could potentially restrict competitors' access to essential competitive inputs.

12. Algorithmic Content Flooding

AI dramatically increases the amount of content that can be produced.

A dominant platform could theoretically generate massive quantities of its own content and thereby make competing human-produced material harder to discover.

The competition issue is particularly significant where the platform controls both:

content production → ranking → recommendation → monetisation.

This creates a potential feedback loop:

AI production → greater content volume → greater platform engagement → more data → better AI → further displacement.

13. Network Effects

AI-generated media platforms may benefit from strong network effects.

More users produce:

→ more data
→ better recommendations
→ more engagement
→ greater advertising revenue
→ greater investment in AI
→ better content generation
→ more users.

Human creators may consequently face increasing difficulty entering markets dominated by large AI-enabled platforms.

14. Case Laws

The following cases are particularly useful by analogy because there are still relatively few final judicial decisions directly determining competition claims arising specifically from AI-generated media displacing human creators. The cases below establish principles concerning digital platforms, data, copyright, technological innovation, market power, exclusion and competition.

Case 1: United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001)

Facts

Microsoft was found to have engaged in exclusionary conduct relating to its operating-system dominance and browser competition.

Principle

The case established important principles concerning:

• monopoly power;
• exclusionary conduct;
• technological innovation;
• platform control; and
• leveraging dominance into adjacent markets.

Relevance to AI-generated media

An AI platform possessing substantial market power could potentially use dominance in one market to disadvantage competing creative services.

For example:

AI model dominance → control over content-generation tools → preferential distribution → foreclosure of independent creators.

Microsoft therefore provides an important framework for analysing technology-driven exclusion.

Case 2: Google Search (Shopping) – European Commission, Case AT.39740 (2017)

Facts

The European Commission found that Google had abused its dominant position by systematically giving prominent placement to its own comparison-shopping service while according competing comparison-shopping services less favourable treatment.

Principle

A dominant platform may create competition concerns when it favours its own downstream service through control over an important distribution platform.

Relevance

The same analytical concern can arise if an AI-powered media platform:

• generates its own content;
• controls search or recommendation;
• ranks its own AI content preferentially; and
• disadvantages human creators.

The important issue is not simply that a company produces AI content, but whether platform control is used to distort competition between the platform's own content and competing creators.

Case 3: United States v. Google LLC, 2024 – Search Distribution Litigation

Facts

The U.S. Department of Justice challenged Google's conduct concerning distribution arrangements and access points for search.

The litigation concerned Google's use of distribution agreements and its position in general search services.

Principle

Competition analysis can examine how control over important distribution channels affects rivals' ability to reach consumers.

Relevance

AI-generated media may increasingly compete through platforms where:

• search;
• recommendation;
• advertising;
• content generation; and
• distribution

are vertically integrated.

If a dominant platform controls both the creative production technology and the audience-access mechanism, foreclosure concerns may become significant.

Case 4: Authors Guild v. Google, Inc., 804 F.3d 202 (2d Cir. 2015)

Facts

Google digitised large numbers of books and created a searchable database.

The litigation principally concerned copyright and fair use rather than antitrust law.

Principle

The case demonstrated the enormous competitive and informational significance of large-scale digitisation and searchable datasets.

The court upheld Google's book-search project under the circumstances presented.

Relevance to AI

Generative AI systems similarly depend upon very large collections of information.

The case is relevant to understanding the distinction between:

• control over information;
• technological transformation;
• copyright interests; and
• competitive advantages arising from large-scale datasets.

It also illustrates why data access can become strategically important in digital markets.

Case 5: FTC v. Facebook, Inc. (Meta), 2021–2025 litigation

Facts

The U.S. Federal Trade Commission challenged Meta's acquisitions and conduct involving social-networking markets.

The litigation focused on alleged maintenance of monopoly power and the competitive significance of acquisitions and platform ecosystems.

Principle

Digital-platform competition may need to consider:

• network effects;
• ecosystem advantages;
• data;
• acquisitions;
• user switching;
• nascent competitors; and
• barriers to entry.

Relevance to AI-generated media

AI companies operating creator platforms may acquire:

• generative-AI startups;
• editing platforms;
• creator marketplaces;
• synthetic-media companies; or
• distribution services.

Such acquisitions may raise concerns where they eliminate emerging competitors or reinforce an integrated ecosystem.

Case 6: Google LLC v. Oracle America, Inc., 593 U.S. 1 (2021)

Facts

The dispute concerned Google's use of Java API material in Android.

The U.S. Supreme Court ultimately held that Google's use constituted fair use under the circumstances.

Principle

The case demonstrates the importance of APIs and interoperability in technology ecosystems.

Relevance to AI-generated media

AI creator ecosystems increasingly depend upon:

• APIs;
• model access;
• plugins;
• interoperability;
• content-management systems; and
• distribution interfaces.

If a dominant AI platform restricts interoperability with competing creative tools, competition authorities may need to examine whether such restrictions protect legitimate technological interests or unnecessarily foreclose rivals.

Case 7: Intel Corp. v. European Commission, Case C-413/14 P (2017)

Facts

The European Commission had found that Intel engaged in conduct involving rebates to major computer manufacturers and retailers.

The litigation addressed the proper assessment of whether rebate practices were capable of foreclosing an equally efficient competitor.

Principle

The case is important for analysing exclusionary conduct by dominant undertakings and the need to examine competitive effects.

Relevance to AI media

A dominant AI platform could potentially offer:

• discounted AI-generation services;
• preferential creator fees;
• bundled access;
• rebates;
• advertising credits; or
• exclusive AI-production contracts.

The competition inquiry would examine whether such arrangements actually restrict effective competition rather than assuming that every discount is unlawful.

Case 8: Bronner v. Mediaprint, Case C-7/97 (1998)

Facts

The case concerned access to a newspaper home-delivery system controlled by a dominant undertaking.

Principle

The Court of Justice applied the stringent conditions associated with the essential-facilities doctrine, particularly concerning refusal of access to infrastructure.

Relevance to AI-generated media

The doctrine may become relevant where an AI or media platform controls infrastructure that competitors genuinely cannot reasonably replicate.

Potential examples could include:

• unique creator-distribution infrastructure;
• indispensable AI datasets;
• specialised AI models;
• essential recommendation systems; or
• critical media-distribution networks.

However, not every successful AI platform constitutes an essential facility. The stringent legal conditions must be satisfied.

15. AI and Collective Bargaining by Human Creators

Human creators may respond collectively to AI substitution.

Actors, writers, musicians, illustrators and other professionals may seek:

• minimum AI compensation;
• consent requirements;
• restrictions on digital replicas;
• licensing arrangements;
• collective bargaining;
• attribution;
• transparency concerning training data.

Competition law must distinguish legitimate collective bargaining from arrangements that unlawfully restrict competition.

At the same time, labour-market competition issues are increasingly relevant because AI may affect the buyer power of firms purchasing creative labour.

16. Monopsony Concerns

Competition law traditionally focuses heavily on sellers and consumers, but AI-generated media may produce labour-market monopsony concerns.

Imagine a market with many creators but only a few major platforms.

If those platforms control access to audiences, they may possess substantial buyer power.

AI may increase this power if platforms can credibly threaten:

“If human creators charge higher prices, we can substitute AI-generated content.”

This could weaken creators' bargaining position.

The competitive inquiry therefore has two dimensions:

Product market: AI content versus human-created content.

Input market: platforms purchasing creative labour from human creators.

17. Creator Data as a Competitive Input

Creators contribute valuable data through:

• engagement;
• audience preferences;
• viewing patterns;
• comments;
• creative styles;
• performance information; and
• behavioural information.

Platforms can use this information to improve recommendation algorithms and AI systems.

This can create a data-feedback advantage.

Human creators therefore may simultaneously be:

1. suppliers;
2. users;
3. sources of data; and
4. competitors of the platform's AI services.

18. Copyright and Competition Law Intersection

AI-generated media raises a major distinction between copyright law and competition law.

Copyright asks:

Who owns or controls the relevant creative work?

Competition law asks:

Does the conduct distort or restrict competitive conditions?

A platform could potentially have lawful access to particular content while still engaging in anticompetitive conduct through exclusionary distribution practices.

Conversely, a copyright dispute does not automatically constitute an antitrust violation.

The two legal regimes therefore operate differently but may overlap.

19. Human Creativity as a Competitive Differentiator

AI does not necessarily eliminate all demand for human creators.

Human creators may continue to compete through:

• originality;
• reputation;
• authenticity;
• emotional connection;
• cultural knowledge;
• artistic identity;
• live performance;
• personal relationships;
• accountability; and
• contractual guarantees concerning provenance.

Competition may therefore evolve from a simple:

human versus machine

model into a:

human + AI versus human + AI

model.

20. Potential Competition Remedies

Where competition authorities identify unlawful conduct, possible remedies could include:

Structural remedies

• divestiture;
• separation of business units;
• restrictions on acquisitions.

Behavioural remedies

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

Data-related remedies

• data access;
• data portability;
• limits on exclusive data arrangements;
• creator consent mechanisms.

Contractual remedies

• restrictions on exclusivity;
• transparent licensing;
• fair access terms.

21. Important Competition-Law Test

A useful analytical framework is:

Step 1 – Define the market

Determine whether the relevant market concerns:

• human creative services;
• AI-generated media;
• media distribution;
• AI infrastructure;
• creator platforms; or
• multiple interconnected markets.

Step 2 – Determine market power

Consider:

• market share;
• network effects;
• data advantages;
• switching costs;
• infrastructure;
• entry barriers;
• interoperability.

Step 3 – Identify conduct

Examine:

• self-preferencing;
• tying;
• bundling;
• exclusive dealing;
• discriminatory access;
• refusal to deal;
• predatory pricing;
• acquisitions;
• algorithmic ranking.

Step 4 – Examine competitive effects

Ask whether the conduct:

• excludes rivals;
• reduces creator choice;
• increases concentration;
• raises barriers to entry;
• reduces innovation;
• harms quality;
• restricts access to audiences.

Step 5 – Consider efficiencies

AI can generate substantial efficiencies:

• lower costs;
• faster production;
• wider consumer choice;
• accessibility;
• personalised content;
• new creative opportunities.

These efficiencies must be distinguished from exclusionary strategies.

22. Competition Displacement Matrix

23. Key Legal Principles from the Cases

The cases collectively demonstrate several principles relevant to AI-generated media:

1. Technological innovation does not automatically excuse exclusionary conduct.
2. Dominant platforms must be examined carefully where they favour their own downstream services.
3. Data can become an important competitive asset.
4. Network effects can strengthen digital-market concentration.
5. Interoperability may be important in technology ecosystems.
6. Refusal of access to genuinely indispensable infrastructure can raise competition concerns.
7. Exclusionary discounts and rebates require effects-based analysis.
8. Digital acquisitions may matter where they eliminate emerging competitive threats.
9. Copyright and competition law address different legal questions.
10. Low-cost AI output can be beneficial competition unless accompanied by exclusionary conduct.

24. Conclusion

AI-generated media creates a fundamental transformation in the competitive structure of creative industries. It can lower production costs, expand access to creative tools and increase consumer choice, while simultaneously creating risks of displacement for human creators.

The principal competition-law concern is therefore not the mere existence of AI-generated content. The critical question is whether firms controlling AI models, datasets, computing infrastructure, creator platforms and distribution channels use that combined power to restrict competition.

The most significant future issues are likely to involve:

• control over training data;
• AI-platform concentration;
• self-preferencing;
• creator-platform dependence;
• algorithmic ranking;
• exclusive licensing;
• AI acquisitions;
• interoperability;
• labour-market monopsony;
• synthetic-content flooding; and
• the relationship between copyright and competition law.