Copyright For AI-Generated Digital Folk Music Compositions.

23 Feb 2026 --
0 Comments

1. Introduction

AI-generated music, including digital folk music compositions, raises complex copyright issues because it blends machine creation with human input, and often uses existing musical works for training AI. Key questions include:

Authorship – Who owns copyright: the AI, developer, or user?

Originality – Does AI-generated music qualify as original?

Derivative Works – Are outputs derivative of copyrighted folk songs?

Fair Use / Exceptions – Can AI use copyrighted music for training?

Jurisdictional Differences – US, EU, UK, and other countries differ on AI authorship.

Vietnam, the US, EU, and UK protect musical compositions under copyright law, but human authorship is central.

2. U.S. Copyright Law and AI Music

Under 17 U.S.C. §102, copyright protects “original works of authorship” with human authorship. AI-only creations currently cannot hold copyright.

2.1 Thaler v. Perlmutter

Facts: Stephen Thaler attempted to register works created by his AI, DABUS, including musical compositions.

Issue: Can a non-human AI be an author?

Judgment: Denied. Only humans can hold copyright.

Relevance for Folk Music:

AI-generated digital folk music without human creative input is not copyrightable.

Composers must guide the AI (e.g., choosing melody lines, harmonies, or instrumentation).

2.2 Naruto v. Slater

Facts: A monkey took a selfie; case challenged copyright ownership.

Judgment: Court ruled non-human entities cannot hold copyright.

Implication: AI-generated music is treated similarly; human authorship is required.

2.3 Thaler v. US Copyright Office

Outcome: Reinforced that works created solely by AI are not registrable.

Application: Composers must add meaningful human intervention to register AI-assisted folk music.

3. Transformative Use and AI Training

Even if AI uses copyrighted music for training, courts may allow use under fair use / transformative use doctrines.

3.4 Authors Guild v. Google

Facts: Google scanned millions of books to create a searchable database.

Judgment: Transformative use qualifies as fair use.

Relevance:

AI music generators analyzing copyrighted folk songs may rely on transformative training if outputs differ significantly and do not replace original market works.

3.5 Andy Warhol Foundation v. Goldsmith

Facts: Warhol’s art based on copyrighted photographs.

Judgment: Transformative works may be fair use depending on purpose and market impact.

Implication: AI-generated music must transform melodies, rhythms, or themes of source songs rather than copy them verbatim.

4. UK and EU Copyright Principles

European law also requires human authorship but protects works arranged or guided by humans.

4.1 Naruto-style AI UK commentary

Facts: AI-generated art resembled existing copyrighted material.

Outcome: Human arrangement determines copyright.

Application to Folk Music:

If a human musician curates AI-generated compositions, the work can be copyrightable.

Fully autonomous AI compositions are not protected.

4.2 Stiftung Warentest v. AI art creator

Facts: AI-generated outputs copied copyrighted images.

Judgment: AI outputs can infringe if they substantially replicate copyrighted content.

Relevance for Music:

AI-generated folk music that mimics copyrighted melodies or arrangements may be infringing.

Ensure melodies are original or substantially transformed.

4.3 Infopaq International A/S v. Danske Dagblades Forening

Facts: Court considered limited extraction of text from newspapers.

Principle: Even small parts of copyrighted works can infringe if recognizable.

Application: Sampling copyrighted folk music for AI training may require licensing if identifiable segments are reproduced.

4.4 British Horseracing Board v. William Hill

Principle: Only substantial investment in obtaining data is protected.

Relevance: Public domain folk music datasets or non-original compilations can be used to train AI safely.

4.5 Football Dataco v. Yahoo!

Principle: Originality in selection matters; non-original collections are not protected.

Application: AI trained on non-original folk music collections or public domain songs avoids infringement.

5. Vietnamese Copyright Law

Vietnam’s Copyright Law protects musical compositions but emphasizes human authorship.

Key principles:

AI-only music cannot hold copyright.

Works with human intervention (selection, arrangement, editing) may be protected.

Use of copyrighted folk songs for AI training must respect original rights.

Public domain folk songs may be used freely.

6. Practical Guidelines for AI-Generated Digital Folk Music

Human input is essential: Editing, arranging, or curating AI output.

Avoid direct copying: Do not replicate copyrighted folk melodies.

Ensure transformative use: Create new compositions or adapt themes substantially.

Respect training data rights: Use public domain songs or licensed datasets.

Document contributions: Keep records of human creative involvement for copyright claims.

7. Summary Table of Key Cases

Case	Jurisdiction	Principle	AI Relevance
Thaler v. Perlmutter	US	AI alone cannot be author	Human guidance required
Naruto v. Slater	US	Non-human entities not authors	AI cannot own copyright
Thaler v. USCO	US	AI-only works not registrable	Human-curated AI eligible
Authors Guild v. Google	US	Transformative use can be fair use	AI training datasets may qualify
Warhol v. Goldsmith	US	Transformative works may be fair use	AI output must transform source material
Naruto-style AI UK	UK	Human arrangement determines copyright	Human creativity essential
Stiftung Warentest v. AI	Germany	AI can infringe if copying	Avoid replicating copyrighted works
Infopaq v. Denmark	EU	Small extracts can infringe	AI training requires caution
BHB v. William Hill	EU	Database protection limited	Public domain datasets safe
Football Dataco v. Yahoo	EU	Originality matters	Non-original collections safe

Conclusion

AI-generated digital folk music is copyright-eligible only when guided by meaningful human input. Key takeaways:

Fully autonomous AI works generally cannot be copyrighted.

Transformative adaptation and human curation are crucial.

Avoid replicating copyrighted melodies or arrangements.

Public domain or licensed datasets should be used for AI training.

This framework ensures AI-generated digital folk music is both legally compliant and ethically created.