Public Domain Contamination Audit for AI Children’s Books
Gary WhittakerThe Second Edition Author System · Article 3 of 5
```
How to Keep a Public-Domain Children’s Book From Accidentally Copying Later Adaptations
Before you generate new AI visuals, separate the original public-domain source from later movies, stage versions, famous lines, costumes, trademarks, and visual signals that do not belong in your second edition.
```Public domain does not usually fail because the creator picked the wrong story.
```It usually fails because the creator did not separate the original source from everything added later.
That is the purpose of the contamination audit.
A public-domain source can give you a legal and creative foundation. But if your prompts, images, cover design, product copy, or character designs pull from later movies, stage versions, famous lines, protected artwork, trademarks, or recognizable brand presentation, the project can become much weaker than it needed to be.
The goal is not fear. The goal is discipline. Before you ask AI to help rebuild the visuals, identify what does not belong.
```Where we are in the Second Edition Author System
```In Article 1, we started with the source record.
The lesson was simple: public domain is the starting point, not the whole permission slip. A creator should document the exact source being used before creating prompts, images, layouts, or product pages.
In Article 2, we moved to rights before redesign.
That article focused on author approval, illustrator rights, publisher records, existing files, AI-input permissions, and the difference between using an old book as a scene map and using old artwork as an AI reference.
Article 3 adds the next layer: separating the public-domain source from later adaptations before the second edition enters visual production.
```What contamination means in a public-domain project
```Adaptation contamination happens when a creator thinks they are using a public-domain source, but the project starts borrowing from later versions that may have their own rights, visual identity, branding, or customer-confusion risk.
It is not always intentional copying. Sometimes it happens because the creator remembers the movie better than the book. Sometimes it happens because a prompt is vague. Sometimes it happens because an AI tool fills in the most familiar cultural version. Sometimes it happens because the product page makes the book look connected to a famous entertainment property.
For beginner authors, the easiest way to understand contamination is this:
If you cannot show whether an element comes from the public-domain source, from your own original work, or from a later adaptation, pause before using it.
Why AI makes contamination easier
```AI tools can move fast, but they do not automatically know which version of a story world you mean.
A prompt like “create a Wizard of Oz children’s book illustration” is not source-specific. It does not tell the tool whether you mean the original book, a film, a stage version, a modern retelling, a fan interpretation, or a general cultural memory of the story.
That is why a contamination audit belongs before the character bible, before the Leonardo workflow, before the page-by-page scene prompts, and before the final image set.
Weak prompt
“Create a Wizard of Oz children’s book illustration.”
This does not define the source, avoid later adaptations, or require an original visual direction.
Cleaner prompt direction
“Create an original children’s storybook illustration inspired by a documented public-domain fantasy source. Use a new visual design, no famous movie costume references, no later adaptation imagery, no recognizable branded elements, no text in the image, and no direct imitation of existing posters or film scenes.”
A better prompt does not replace legal review. It simply gives the creator a cleaner production boundary.
```The three-bucket method
```The simplest way to run a contamination audit is to divide the project into three buckets: public-domain source, original additions, and do not use.
| Public-Domain Source | Original Additions | Do Not Use |
|---|---|---|
| Elements documented from the original public-domain source. | New creative choices made by the current author or second-edition team. | Elements from later adaptations, confusing brand signals, unclear modern rights, or unapproved prior artwork. |
| Source-supported characters, places, general story world, and descriptions confirmed in the original book. | New scenes, new dialogue, new page layout, new cover direction, new visual bible, new product positioning. | Movie-only costume details, famous film lines, later stage imagery, official-looking logos, film-poster composition, and prior illustrator art used as AI input without rights. |
If an element does not clearly fit into one of these buckets, do not use it until it has been checked.
```Oz examples creators can understand
```Oz is a strong teaching example because many people know the world, but not everyone knows which details come from the public-domain book and which details became famous through later adaptations.
This does not mean beginner authors should avoid public domain. It means they should check details before using them.
| Area | Contamination Risk | Cleaner Approach |
|---|---|---|
| Shoes and costumes | Using famous later visual details without checking whether they come from the original source. | Verify details in the original book and create a fresh visual design for the new edition. |
| Famous lines | Assuming a quote is from the public-domain book because it is culturally familiar. | Use the author’s own dialogue or verify the line in the source before using it. |
| Witch presentation | Copying later visual framing, makeup, costume, personality treatment, or stage identity. | Design an original version that serves the author’s story and does not imitate famous adaptation cues. |
| Cover design | Making the book look official, film-like, stage-branded, or visually connected to a later entertainment property. | Create new typography, new composition, new color rules, and clear independent product language. |
| AI prompt language | Prompting with “like the famous movie” or relying on vague cultural memory. | Prompt from the source record, original creative brief, and specific “do not use” list. |
| Product page copy | Implying affiliation, endorsement, or official connection. | State that the work is independent and inspired by public-domain source material. |
Applying the audit to the second-edition case study
```For the Oz-inspired second-edition case study, the rebuild should separate four things.
The documented Oz source foundation.
The original second-edition story being preserved and improved.
Existing layout, illustrations, credits, and publishing records that need rights review.
Movie, stage, brand, and visual material that should not guide the redesign.
The case study should use the public-domain source as context, keep the author’s story as the foundation, treat the prior edition as a rights-sensitive scene map, and avoid later adaptation cues when building the new visual identity.
That is how the second edition can become cleaner, stronger, and more original instead of simply looking like an AI repaint of something that already existed.
```The Public-Domain Contamination Audit Checklist
```Use this checklist before writing final prompts, generating the full image set, creating the cover, writing the Shopify product page, or uploading a book to KDP.
1. Source questions
- What exact public-domain source am I using?
- Who created the original source?
- When was it published?
- What elements am I taking from that source?
- Have I verified those elements in the source itself?
2. Adaptation questions
- What later adaptations do people associate with this story?
- What movie, stage, TV, book, or brand versions should I avoid?
- Are there famous lines people expect from later versions?
- Are there costume, color, or character details that belong to later versions?
- Are there visual designs that feel too close to a known adaptation?
3. AI prompt questions
- Does my prompt identify the source clearly?
- Does my prompt tell the AI what to avoid?
- Does my prompt ask for an original visual design?
- Am I using any existing artwork as a reference?
- Do I have written rights for every image I upload as reference?
- Did I review outputs for contamination before selecting final art?
4. Product page questions
- Does the title create confusion?
- Does the cover look like a later adaptation?
- Does the product description imply official connection?
- Does the disclaimer explain the work is independent?
- Does the page explain the original contribution?
5. Approval questions
- Has the author approved the source-safe direction?
- Has the illustrator rights question been handled?
- Has the new visual direction been approved before full production?
- Has the final product copy been reviewed before publication?
Sample audit table for an Oz-inspired second edition
```This is a beginner example. It does not replace legal review, but it gives the creator a practical working boundary before AI production begins.
| Keep | Rebuild | Avoid |
|---|---|---|
| The author’s original story foundation. | The second-edition layout, cover, typography, visual bible, and production files. | Famous later adaptation visuals, official-looking branding, and movie/stage-specific cues. |
| Source-supported public-domain context. | New character sheets, new scene prompts, new image folders, and new prompt logs. | Prior illustrator artwork as AI input unless written permission allows that use. |
| Clear independent public-domain inspiration language. | The Shopify product page, case-study language, author page, and launch copy. | Product copy that implies affiliation, endorsement, or official connection. |
Careful language for public copy
```Public-domain projects need careful public language. The right words protect reader trust and reduce confusion.
Instead of: “Oz is free to use.”
Use: “This independent project is inspired by public-domain source material from the original Oz stories.”
Instead of: “AI updated the old illustrations.”
Use: “The second edition uses a new visual direction created through a documented AI-assisted workflow.”
Instead of: “This is the new version of Oz.”
Use: “This is an independent children’s book inspired by public-domain source material.”
Instead of: “Based on the classic movie.”
Use: “Built from documented public-domain source research and original creative development.”
Build your contamination audit
```Before moving into the AI visual system, answer these questions in your project folder.
1. My public-domain source is: ______________________________
2. The later adaptations people may confuse with it are: ______________________________
3. The details I must verify before using are: ______________________________
4. The visual signals I must avoid are: ______________________________
5. The phrases or famous lines I must avoid are: ______________________________
6. My original additions are: ______________________________
7. My new visual direction will be: ______________________________
8. My AI prompt avoid-list is: ______________________________
9. My product page disclaimer will say: ______________________________
10. I need legal or professional review for: ______________________________
If you cannot complete these answers, the project is not ready for full image production.
```Why this audit comes before the AI visual system
```Article 4 will move into the creative production system: character bible, scene map, prompt structure, image folders, Leonardo workflow, and page-by-page visual planning.
But that workflow should not begin until the source boundaries are clear.
Without a contamination audit, the character bible can accidentally bake in risky details. The cover can accidentally imitate a later adaptation. The Shopify copy can accidentally imply an official connection. The AI prompts can accidentally guide the tool toward protected visual memory.
The audit prevents that before it becomes expensive to fix.
```What to do before moving to Article 4
```Before creating character sheets or AI image prompts, complete this pass.
- Confirm the public-domain source record from Article 1.
- Confirm the rights review from Article 2.
- List the later adaptations that could contaminate the project.
- Write a “do not use” list for prompts, visuals, cover design, and product copy.
- Write the original creative direction for the second edition.
- Review whether existing artwork will be used only as a scene map or as AI input.
- Remove any visual or wording choices that could imply official connection.
- Save the audit before building the visual bible.
Coming next: the AI visual system
```Article 1 created the source record.
Article 2 reviewed rights before redesign.
Article 3 separated the public-domain source from later adaptation contamination.
Article 4 will show how to turn those records into a visual bible, character sheets, scene prompts, and a Leonardo AI workflow for a second-edition children’s book.
```Responsible creator note
```This article is educational. It is not legal, tax, or publishing-platform approval advice.
If your project involves a known story world, an existing illustrator, a prior publisher, an existing ISBN, a trademark question, a publishing contract, AI-generated artwork, or copyright registration, speak with a qualified professional before you publish or sell the new edition.
```Source links for readers
```These are starting points for creators who want to check the source and platform rules behind this article.
Final Word
```Before you prompt the new visuals, audit what does not belong.
Public domain can help you begin, but the finished book still needs source discipline, original expression, and careful separation from later adaptations. If you want your second edition to stand on its own, do not build from memory. Build from a documented source, a clear avoid list, and a new creative direction.
```