Copyright Issues In AI-Generated Therapeutic Bedtime Narrations.

28 Feb 2026 --
0 Comments

I. Core Copyright Issues in AI-Generated Therapeutic Narrations

1. Authorship and Human Creativity

Bedtime stories

Guided meditation scripts

Therapeutic affirmations

The central question is whether a human author contributed sufficient creativity. Pure AI-generated content without human input is generally not copyrightable.

2. Training Data Concerns

AI models are trained on:

Public domain stories

Licensed content

Copyrighted books or scripts

If copyrighted works are used without authorization, there is a risk of infringement. Exceptions may exist under:

Fair use (U.S.)

Text and data mining exceptions (EU)

3. Substantial Similarity & Derivative Works

If AI-generated content closely mimics:

Existing copyrighted bedtime stories

Meditation scripts

Audio recordings

It can lead to infringement claims. The substantial similarity test is used to evaluate these cases.

4. Idea–Expression Dichotomy

Not protected: General narrative ideas, sleep guidance techniques, therapeutic methods.

Protected: Specific story text, unique phrases, metaphors, arrangement of sentences, audio performance.

5. Audio Recordings & Performance Rights

For narrated content, the AI-generated audio may raise additional questions:

Is the AI’s voice a protected performance?

Can imitation of a human voice infringe rights of publicity or sound recording copyright?

Courts are still developing standards for AI audio output.

II. Leading Case Laws Relevant to AI-Generated Therapeutic Narrations

1. Feist Publications, Inc. v. Rural Telephone Service Co.

Facts: Feist copied phone listings from Rural’s directory.
Holding: Facts themselves are not copyrightable; only original selection/arrangement is.
Principle: Minimal creativity is required for compilation.
Application: AI bedtime narrations can freely use general story ideas, plot structures, or common therapeutic themes, but creative expression in storytelling is protected.

2. Authors Guild v. Google, Inc.

Facts: Google digitized millions of books to create a searchable database.
Holding: Court held digitization was fair use because it was transformative.
Principle: Transformative use is key; copying for indexing/search may be permissible.
Application: Training AI on copyrighted bedtime stories could be allowed if used transformatively to generate new, original therapeutic scripts, not verbatim copies.

3. Naruto v. Slater

Facts: A monkey took selfies with a photographer’s camera.
Holding: Animals cannot hold copyright.
Principle: Only humans can be authors.
Application: Pure AI-generated stories without human input cannot be copyrighted, highlighting the need for human contribution in therapeutic narration creation.

4. Thaler v. Perlmutter

Facts: Stephen Thaler sought copyright for AI-created artwork.
Holding: Denied because there was no human author.
Principle: Human authorship is mandatory for copyright.
Application: For AI-generated bedtime narrations, human involvement—editing, selection, or creative guidance—is essential for copyright protection.

5. Andy Warhol Foundation v. Goldsmith

Facts: Warhol created stylized art from Goldsmith’s photograph of Prince.
Holding: Commercial use of copyrighted image without sufficient transformation was infringement.
Principle: Transformative use is limited; market substitution matters.
Application: If AI bedtime scripts closely mimic commercially published therapeutic stories, fair use may not protect them if they substitute the market for the original works.

6. Oracle America, Inc. v. Google LLC

Facts: Google copied portions of Oracle’s Java API for Android.
Holding: Copying was fair use, but functional elements have thinner protection.
Principle: Functional code or frameworks are protected differently from expressive content.
Application: AI frameworks generating scripts may use general storytelling structures (functional templates) without infringing copyright.

7. Infopaq International A/S v. Danske Dagblades Forening

Facts: A company copied 11-word newspaper snippets.
Holding: Even short excerpts may infringe if they reflect originality.
Principle: Small segments can be protected if expressive.
Application: Even a few lines of text in a bedtime narration can be copyrighted if they show original expression, so AI output must avoid verbatim replication of copyrighted stories.

8. SAS Institute Inc. v. World Programming Ltd.

Facts: Competitor replicated SAS software functionality.
Holding: Functionality and programming language are not copyrightable.
Principle: Ideas, methods, procedures are free to use.
Application: Storytelling frameworks, meditation techniques, or narrative sequences (functional aspects) can be used freely; only the expressive content is protected.

III. Key Legal Themes for AI-Generated Bedtime Narrations

Human authorship is required – AI alone cannot hold copyright.

Ideas vs expression – General story structures or sleep guidance are free to use; unique text/audio expression is protected.

Transformative use is limited – Market substitution and commercial intent are scrutinized.

Small excerpts can infringe – Even a few lines of creative text matter.

Functional frameworks are safer – Story templates, meditation frameworks, or narrative models are not copyrightable.

IV. Practical Guidance for Developers

Ensure human creative contribution (editing, selecting, or curating AI output).

Avoid direct copying of copyrighted stories or scripts.

Use licensed or public domain content for training.

Document AI prompt and output modifications to establish authorship.

Separate functional narrative frameworks from expressive, creative text.

Consider market impact to avoid potential commercial infringement.

V. Conclusion

AI-generated therapeutic bedtime narrations face copyright risks primarily in:

Textual originality

Audio expression

Derivative works

Courts consistently emphasize human authorship, originality, and transformative use limits. Cases like Thaler, Naruto, Warhol, and Infopaq show that AI-generated works without human creative input are not protected, while functional frameworks are free to use.