Ownership Of AI-Curated Digital Botanical Databases

16 Feb 2026 --
0 Comments

I. FOUNDATIONAL LEGAL QUESTIONS

When AI curates a botanical database (e.g., plant taxonomy, genomic data, herbarium images, ecological metadata), the ownership issues typically include:

Who owns the raw botanical data?

Is the database protected as a copyrightable compilation?

Does AI-generated structure qualify for copyright?

Is there sui generis database protection (EU)?

Who owns AI outputs — developer, user, or no one?

Are plant-related IP regimes (patents/PBR) implicated?

What role do contracts and licenses play?

II. KEY COPYRIGHT PRINCIPLES GOVERNING DATABASES

1. Originality Requirement

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

Facts: Rural Telephone created a white pages directory. Feist copied listings.
Issue: Can factual compilations be copyrighted?
Held: Facts are not copyrightable. Only original selection or arrangement is protected.

Legal Significance:

Botanical facts (plant names, taxonomy, DNA sequences) are not copyrightable.

However, creative selection, categorization, annotations, taxonomic grouping, metadata structure may qualify.

AI-curated classification must show human intellectual contribution.

If AI independently organizes plant species without human creative control, copyright protection becomes doubtful.

2. Sweat of the Brow Rejected

➤ University of London Press Ltd v University Tutorial Press Ltd

Principle: Originality requires skill, labor, and judgment — not mere copying.

Applied to botanical databases:

Simply investing money to digitize herbarium records does not create copyright.

Intellectual effort in curating ecological correlations or evolutionary clustering may qualify.

3. Human Authorship Requirement in AI Context

➤ Naruto v. Slater

Facts: A monkey took a selfie; dispute over copyright.
Held: Non-humans cannot hold copyright.

Implication for AI:

If an AI system autonomously curates botanical classifications:

The AI cannot be the author.

The database must have human intellectual input to be protected.

This principle strongly affects ownership of AI-curated botanical datasets.

➤ Thaler v. Perlmutter

Facts: Stephen Thaler sought copyright registration for AI-generated artwork.
Held: Human authorship is required under US copyright law.

Application:

If an AI autonomously structures plant phylogenetic databases:

Pure AI-generated structuring = no copyright

Human-guided AI (prompted, reviewed, curated) = possible protection

Thus, ownership depends on degree of human creative control.

III. DATABASE-SPECIFIC PROTECTION (EU)

4. Sui Generis Database Right

➤ British Horseracing Board Ltd v William Hill Organization Ltd

Issue: What counts as "substantial investment" in database creation?
Held: Investment must relate to obtaining, verifying, or presenting existing data, not creating the data itself.

Application to Botanical Databases:

If an organization:

Invests heavily in collecting plant field data

Digitizes herbarium specimens

Verifies taxonomic classifications

They may claim database rights in the EU, even if copyright fails.

Important distinction:

Investment in generating plant data ≠ database right

Investment in organizing/verifying data = protected

➤ Football Dataco Ltd v Yahoo! UK Ltd

Issue: Does intellectual creation matter for database copyright?
Held: Database must reflect author's intellectual creation.

Applied here:
If AI automatically clusters species based purely on algorithms:

Likely insufficient
If botanists determine weighting criteria, classification thresholds, ecological mapping:

Stronger copyright claim

IV. COMPILATION AND ARRANGEMENT PROTECTION

5. Creative Selection and Arrangement

➤ Key Publications Inc v Chinatown Today Publishing Enterprises Inc

Facts: Directory of Chinese-American businesses.
Held: Selective inclusion based on subjective criteria qualifies for copyright.

Application:
If botanical database:

Includes endangered species only

Groups plants by medicinal significance

Organizes flora by climate vulnerability

Then selection may qualify as protectable compilation.

6. Modicum of Creativity

➤ CCC Information Services Inc v Maclean Hunter Market Reports Inc

Court upheld protection in valuation guides because of subjective judgment.

Applied:
AI-curated plant valuation for commercial crops:

If human experts define classification metrics,

Then resulting database likely protected.

V. AUSTRALIAN PERSPECTIVE

➤ IceTV Pty Ltd v Nine Network Australia Pty Ltd

Held: Mere labor insufficient; must show independent intellectual effort.

Implication:
Automated botanical aggregation without human intellectual input likely unprotected in Australia.

VI. INTERACTION WITH PLANT-SPECIFIC IP REGIMES

AI-curated botanical databases may intersect with:

1. Plant Variety Protection (PVP)

Under regimes like:

UPOV Convention

Plant Breeders' Rights Acts

These protect new plant varieties, not databases.

2. Patent Law

Genetically modified plant traits may be patentable (e.g., biotech crops).

However:

Raw botanical data ≠ patentable.

Database structure may not qualify unless technical innovation exists.

VII. WHO OWNS AI-CURATED BOTANICAL DATABASES?

Ownership depends on scenario:

Scenario	Likely Owner
Human curates with AI assistance	Human/Institution
Fully autonomous AI structuring	Possibly no copyright
EU substantial investment	Database producer
Contractually assigned rights	As per agreement
Public funding research	Often institution or public domain

VIII. CONTRACTUAL CONTROL

Even where copyright fails:

Data licenses

API agreements

Research contracts

can create enforceable ownership-like control.

This is increasingly how botanical AI platforms protect their datasets.

IX. KEY DOCTRINAL THEMES FROM THE CASES

Facts are free – Feist

Human authorship required – Thaler

Skill and judgment required – IceTV

Selection and arrangement protectable – Key Publications

Investment protects databases (EU) – British Horseracing

Original intellectual creation needed – Football Dataco

X. EMERGING UNRESOLVED ISSUES

Can “AI-assisted human creativity” satisfy originality?

How much human intervention is sufficient?

Will jurisdictions create AI-specific database protection?

Should ecological public-good data be privately owned?

Courts globally are still developing principles.

XI. CONCLUSION

Ownership of AI-curated digital botanical databases depends on:

Degree of human intellectual contribution

Jurisdiction (US vs EU difference significant)

Nature of investment

Contractual arrangements

Whether protection claimed is copyright or database right

The dominant global rule remains:

No human intellectual creativity → No copyright ownership.

However, where humans guide AI in selecting, verifying, and organizing botanical data, ownership of the database as a compilation is legally sustainable.