Copyright Issues In AI-Generated Scientific Illustrations For Local Species.
I. Core Copyright Issues in AI-Generated Scientific Illustrations
1. Authorship and Human Originality Requirement
Most copyright systems (U.S., U.K., EU, India, Australia, etc.) require human authorship.
If an AI tool generates a scientific illustration of, for example, a rare frog species found in a particular region, key legal questions arise:
Who is the author?
Is there sufficient human creative input?
Is the output copyrightable at all?
If the illustration is entirely machine-generated without meaningful human creative control, many jurisdictions will deny copyright protection.
2. Idea–Expression Dichotomy in Scientific Illustration
Species themselves are facts of nature, not copyrightable.
You cannot copyright:
The anatomy of a plant
The morphology of an insect
The scientific facts about a species
However, you can copyright:
The artistic style
The selection of perspective
Shading, composition, color interpretation
Label arrangement and creative presentation
Thus, AI outputs risk infringement if they replicate expressive elements of prior scientific illustrations.
3. Training Data and Derivative Works
AI systems are trained on massive datasets, potentially including:
Field guides
Museum drawings
Taxonomy monographs
Conservation posters
If the AI reproduces distinctive expressive elements of an earlier illustrator’s work, it may constitute:
Derivative work
Substantial similarity infringement
Style appropriation (though style alone is not protected)
4. Database and Compilation Issues
Local species datasets compiled by universities or biodiversity boards may be protected as:
Compilations
Databases
Institutional publications
Even if species themselves are not protected, the selection and arrangement might be.
5. Moral Rights Concerns
In countries recognizing strong moral rights (e.g., France, India):
Attribution rights
Integrity rights
If AI outputs mimic or distort a known illustrator’s style, moral rights may be implicated.
II. Key Case Laws Relevant to AI-Generated Scientific Illustrations
Below are eight major cases, explained in depth.
1. Burrow-Giles Lithographic Co. v. Sarony
Background:
This case involved a photograph of Oscar Wilde taken by Napoleon Sarony.
Legal Issue:
Was a photograph protected by copyright?
Holding:
Yes. The Supreme Court held that photographs are protected when they involve creative choices by a human author.
Relevance to AI:
This case establishes:
Copyright requires human intellectual conception.
Mechanical processes do not negate protection if human creativity exists.
Applied to AI:
If a scientist meaningfully directs the AI (angle, anatomical emphasis, lighting, structure), protection may exist.
If the output is autonomous, protection may fail.
2. Feist Publications, Inc. v. Rural Telephone Service Co.
Background:
Rural Telephone published a phone directory. Feist copied listings.
Legal Issue:
Are factual compilations protected?
Holding:
Facts are not protected; only original selection and arrangement are.
Relevance:
Species morphology = facts.
Scientific reality = facts.
AI illustration of:
Exact anatomical representation
Purely factual diagrams
May lack originality unless expressive creativity is present.
This case is critical for:
Biodiversity databases
Species catalogs
Scientific plate collections
3. Bridgeman Art Library v. Corel Corp.
Background:
Corel reproduced exact photographic copies of public domain paintings.
Holding:
Exact reproductions of public domain works lack originality.
Relevance:
If AI recreates:
Exact museum plates
Historic taxonomic illustrations
Public domain natural history drawings
And adds no creative variation → no new copyright arises.
This is important where AI “restores” or “upscales” classic biodiversity illustrations.
4. Naruto v. Slater
Background:
A macaque took selfies using a photographer’s camera.
Legal Issue:
Can a non-human own copyright?
Holding:
No. Only humans can be authors under U.S. copyright law.
Relevance:
AI is not a legal author.
Thus:
Fully autonomous AI-generated scientific drawings may lack copyright protection.
Ownership claims must rely on human involvement.
This case is foundational in AI copyright debates.
5. Thaler v. Perlmutter
Background:
Stephen Thaler attempted to register artwork created by an AI system called “Creativity Machine.”
Holding:
The court reaffirmed that human authorship is mandatory.
Importance:
This is the most direct modern case concerning AI authorship.
Applied to scientific illustration:
If a biodiversity researcher uses AI without meaningful creative intervention, copyright protection may be denied.
6. Andersen v. Stability AI Ltd.
Background:
Artists sued AI image generators for training on copyrighted artworks.
Allegations:
Unauthorized training
Style imitation
Derivative outputs
Current Legal Relevance:
Though ongoing, it raises critical questions:
Is AI training infringing?
Is style mimicry unlawful?
Are outputs derivative?
For scientific illustrators:
If AI replicates a well-known botanical illustrator’s unique style, legal risks arise.
7. Kelly v. Arriba Soft Corp.
Background:
Search engine used thumbnail versions of photographs.
Holding:
Thumbnail use was transformative and fair use.
Relevance:
AI training may be analogized to transformative indexing.
Defense argument:
Training AI on biodiversity images may be transformative and non-expressive copying.
However:
Output reproduction of substantial similarity would negate that defense.
8. American Geophysical Union v. Texaco Inc.
Background:
Texaco photocopied scientific journal articles for research.
Holding:
Not fair use; copying harmed licensing market.
Relevance:
If AI companies copy scientific field guides without license, courts may find market harm.
Especially relevant for:
Local ecological field guides
Commercial wildlife atlases
III. Additional Legal Doctrines Relevant to Local Species Illustrations
A. Merger Doctrine
When idea and expression merge (e.g., only one way to draw a specific anatomical cross-section), protection may be thin.
B. Scenes a Faire Doctrine
Standard scientific conventions (labeled arrows, dorsal view, scale bar) are not protectable.
C. Substantial Similarity Test
Courts compare:
Total concept and feel
Protected expressive elements
Quantitative and qualitative copying
IV. Special Issues in Local Species Context
1. Indigenous Knowledge
Some local species illustrations incorporate:
Traditional ecological knowledge
Tribal symbolism
Cultural expression
These may implicate:
Cultural intellectual property
Moral rights
Traditional knowledge protections (outside strict copyright)
2. Government Publications
If AI trains on:
Government biodiversity reports
Public domain species databases
Legal outcomes depend on jurisdiction:
U.S. federal works = public domain
Other countries = sometimes protected
3. Conservation and Academic Publishing Risks
Journals may reject AI-generated figures if:
Copyright ownership unclear
Licensing ambiguous
Data provenance uncertain
V. Practical Risk Scenarios
Scenario 1:
AI reproduces a rare orchid drawing nearly identical to a known botanical plate → likely infringement.
Scenario 2:
AI produces a scientifically accurate but stylistically novel frog illustration → likely copyrightable if sufficient human input.
Scenario 3:
AI mimics a famous wildlife illustrator’s signature watercolor technique → legal gray area (style not protected, but substantial similarity may be).
VI. Comparative Jurisdictional Overview
| Jurisdiction | Human Authorship Required? | AI Protection? |
|---|---|---|
| U.S. | Yes | No (without human input) |
| U.K. | Computer-generated works allowed (author = arranger) | Limited |
| EU | Strong human originality requirement | Generally no |
| India | Human authorship required | Likely no autonomous AI protection |
VII. Key Legal Tensions Going Forward
Ownership vs. Public Domain
Innovation vs. Artist Protection
Scientific Accuracy vs. Creative Expression
Fair Use vs. Licensing Markets
Indigenous Biodiversity Rights
VIII. Conclusion
AI-generated scientific illustrations of local species sit at the intersection of:
Human authorship doctrine
Factual representation limits
Derivative work analysis
Training data legality
Fair use doctrine
The most influential cases shaping this area include:
Burrow-Giles Lithographic Co. v. Sarony
Feist Publications, Inc. v. Rural Telephone Service Co.
Bridgeman Art Library v. Corel Corp.
Naruto v. Slater
Thaler v. Perlmutter
Andersen v. Stability AI Ltd.
Kelly v. Arriba Soft Corp.
American Geophysical Union v. Texaco Inc.
Together, they create the doctrinal framework that courts will apply to AI-generated biodiversity and scientific illustrations.
RELATED Blog
comments