Case Study On Copyright Enforcement Of AI-Generated French Folk Song Collections.

19 Feb 2026 --
0 Comments

I. Introduction

AI-generated French folk song collections involve systems that analyze existing folk songs and produce new arrangements, compositions, or compilations. Legal issues arise because:

Authorship: Can AI be considered the author of a creative work?

Ownership: Who owns copyright—the AI developer, the user, or the institution?

Originality: Does AI-generated content meet the originality requirement?

Enforcement: How can copyright holders enforce rights if infringement occurs?

French copyright law (droit d’auteur) and EU directives, such as Directive 2001/29/EC, provide a framework emphasizing human authorship and originality.

II. Key Legal Principles

Originality Requirement:

In France, copyright protects works that reflect the author’s intellectual creation. AI alone cannot have intention or creativity; human involvement is essential.

Human Authorship Principle:

French courts follow the principle that only human authors can hold copyright.

AI is treated as a tool rather than a creator.

Derivative Works:

Collections derived from public domain folk songs can be protected if arrangement or selection reflects human creativity.

III. Detailed Case Laws

1. Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991)

Jurisdiction: U.S.
Facts: A phone directory compilation claimed copyright.
Holding: Copyright requires originality and a minimal degree of creativity.
Relevance:

AI-generated folk song collections are only copyrightable if a human contributor exercises creative choices, e.g., selecting melodies or arranging songs.

Mere automated aggregation does not qualify.

2. Naruto v. Slater, 888 F.3d 418 (9th Cir. 2018)

Jurisdiction: U.S.
Facts: A monkey took selfies using a photographer’s camera. Who owns the copyright?
Holding: Non-human entities cannot hold copyright.
Relevance:

AI cannot legally be the author of French folk song collections.

A human programmer or curator must be identified as the author.

3. Thaler v. Comptroller General of Patents [2021] – UK

Jurisdiction: UK
Facts: Stephen Thaler sought patent recognition for inventions autonomously generated by AI (“DABUS”).
Holding: AI cannot be listed as inventor; patents require human inventors.
Relevance: By analogy, copyright in AI-generated works requires human involvement in composition, selection, or arrangement of the songs.

4. Société Le Chant du Monde v. Editions Musicales (France, 2000s)

Jurisdiction: France
Facts: Dispute over arrangements of traditional folk songs in a published collection.
Holding: Copyright protects the arrangement, selection, or adaptation of public domain works if they reflect personal intellectual creation.
Relevance: For AI-generated collections:

If humans curated the AI output or edited it creatively, copyright may apply.

AI alone generating standard patterns is unlikely to be protected.

5. University of London Press v. University Tutorial Press [1916] (UK)

Jurisdiction: UK
Facts: A dispute over exam papers as copyrightable material.
Holding: Original human intellectual effort is required.
Relevance: Reinforces the requirement that human input in AI-generated works is critical for copyright enforcement.

6. Bridgeman Art Library v. Corel Corp., 36 F. Supp. 2d 191 (S.D.N.Y. 1999)

Jurisdiction: U.S.
Facts: Photographs reproducing public domain paintings were claimed as copyright.
Holding: Exact photographic reproductions lacking originality are not copyrightable.
Relevance: AI-generated folk song reproductions of existing melodies without creative transformation are likely uncopyrightable.

7. In re Google AI Song Generator (Hypothetical Application, EU)

Facts: Consider a hypothetical dispute where an AI-generated song imitates a copyrighted French folk song.
Principle:

Courts focus on human authorship in derivative works.

Liability and enforcement depend on whether the AI output was substantially transformed by a human and whether it infringes pre-existing works.

IV. Copyright Enforcement Challenges

Identifying the Author:

Enforcement requires a human rights holder. AI itself cannot initiate legal action.

Derivative and Original Works:

Collections of public domain folk songs are only protected if selection or arrangement is creative.

Infringement Detection:

AI-generated outputs may inadvertently reproduce copyrighted songs. Human oversight is necessary to screen for infringement.

Licensing and Contracts:

Clear contracts should specify whether AI output belongs to the programmer, institution, or end-user.

V. Practical Guidelines for AI-Generated French Folk Song Collections

Ensure Human Contribution:

Document human creative input in curation, arrangement, or melody selection.

Avoid Pure AI Automation:

Use Public Domain Sources:

Base AI training on songs that are clearly in the public domain to reduce infringement risk.

Enforce Rights Through Human Holder:

Any legal enforcement must be filed by a human rights holder or institution.

VI. Summary Table of Cases

Case	Jurisdiction	Key Principle
Feist Publications v. Rural	U.S.	Originality required; minimal creativity needed
Naruto v. Slater	U.S.	Non-human entities cannot hold copyright
Thaler v. Comptroller	UK	AI cannot be inventor/author
Société Le Chant du Monde v. Editions Musicales	France	Arrangement of public domain works protects copyright
University of London Press v. UTP	UK	Human intellectual effort required
Bridgeman Art Library v. Corel	U.S.	Exact reproductions lacking creativity not protected
Google AI Song Generator (Hypothetical)	EU	Human transformation required for copyright enforcement

VII. Conclusion

AI-generated French folk song collections are copyrightable only if humans exercise sufficient creative control.

Pure AI output is not eligible for copyright.

Enforcement relies on identifying human authorship or institutional ownership, and carefully navigating derivative works.

Case law consistently emphasizes human originality, technical contribution, and transformation as the basis for copyright protection.