Ownership Of Algorithmically Curated Digital Biodiversity Databases
1. Introduction: Ownership of Digital Biodiversity Databases
A digital biodiversity database may include species records, DNA sequences, ecological metadata, images, and observations. When algorithms automatically curate, classify, or analyze data:
Questions arise: Who owns the database?
Is it the human creator of the algorithm, the data collector, or the organization hosting the system?
Can an AI alone be an owner, or are ownership rights inherently human/legal?
Ownership is usually examined through three lenses:
Copyright – Protects original selection or arrangement of data, not facts.
Database Rights (Europe) – Sui generis rights protect substantial investment in obtaining, verifying, or presenting data.
Patent/Trade Secret – May apply to underlying algorithms but not the raw data.
2. Key Legal Challenges
Originality vs. Automation – Copyright requires human authorship. Fully automated AI curation may lack traditional authorship.
Facts vs. Selection/Arrangement – Facts themselves are not copyrightable; only the original selection, organization, and presentation can be protected.
Ownership of AI Output – Courts are increasingly skeptical about granting copyright to AI-generated works.
Collaborative Data Ownership – Biodiversity databases often combine data from multiple sources, complicating ownership claims.
3. Important Case Laws
(A) Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991)
Facts: Feist copied a white pages directory compiled by Rural Telephone Service. Rural claimed copyright on the directory.
Ruling: The Supreme Court held that facts are not copyrightable, and only original selection and arrangement can be protected.
Relevance: Algorithmically curated biodiversity databases cannot claim copyright on raw species data, only on the unique structure, taxonomy, or presentation created by humans.
(B) Naruto v. Slater (“Monkey Selfie”), 888 F.3d 418 (9th Cir. 2018)
Facts: A monkey took a selfie using a photographer’s camera. The photographer claimed copyright.
Ruling: Courts held that non-human authors cannot hold copyright; only humans can.
Relevance: If an AI algorithm autonomously curates a biodiversity database without human input, the AI cannot own copyright, reinforcing the need for human authorship.
(C) Bridgeman Art Library v. Corel Corp., 36 F. Supp. 2d 191 (S.D.N.Y. 1999)
Facts: Corel copied high-quality photographs of public domain artworks.
Ruling: Exact photographic reproductions of public domain works do not create copyright, as there is no originality.
Relevance: Automated digital scans or raw images of species, even curated algorithmically, may not be copyrightable unless creative modifications or organization exist.
(D) International Case: British Horseracing Board Ltd v. William Hill Organization Ltd [2001] UKHL 47 (UK)
Facts: The BHB claimed database rights over horse racing results against a betting company that copied them.
Ruling: The House of Lords held that a substantial investment in creating a database can grant sui generis database rights, even if the underlying data is factual.
Relevance: In Europe, a biodiversity organization investing in gathering, verifying, and maintaining species data may have legal protection over the database structure, even if curated by AI.
(E) CJEU Case: Fixtures Marketing Ltd v. Svenska Spel AB, C-444/02 (EU)
Facts: Examined whether data compilation could qualify for database rights under the EU Database Directive.
Ruling: Database rights exist if there is a substantial investment in obtaining, verifying, or presenting data.
Relevance: Digital biodiversity databases with algorithmic curation could be protected if the organization demonstrates significant investment in maintaining and curating data.
(F) Naruto v. Slater Influence on AI Output Ownership (Reiterated)
Emphasizes that fully autonomous AI output without human selection may lack copyright.
For database ownership, human involvement in algorithm design or data curation is critical.
4. Practical Implications for Ownership
Human-in-the-Loop Principle – Ensure human authorship or supervision of algorithmic curation.
Contracts & Licensing – Ownership may be governed by agreements between contributors, AI developers, and institutions.
Database Rights (Europe) – Highlight investment in data collection, verification, and presentation.
Copyright vs. Raw Data – Only creative selection, arrangement, or presentation can be copyrighted.
5. Summary Table of Key Cases
| Case | Key Point | Relevance to Biodiversity Databases |
|---|---|---|
| Feist v. Rural | Facts not copyrightable; selection/arrangement must be original | Only curated/arranged data by humans protected |
| Naruto v. Slater | Non-human authorship cannot hold copyright | AI alone cannot own database copyright |
| Bridgeman v. Corel | Exact reproductions of public domain works not protected | Raw images or species data not copyrightable |
| BHB v. William Hill (UK) | Substantial investment in data grants database rights | Database rights can protect AI-curated collections |
| Fixtures Marketing (EU) | Substantial effort in compiling data protects database | Investment in curation may be protected under EU law |
✅ Conclusion:
Ownership of algorithmically curated digital biodiversity databases depends on human involvement and legal framework:
U.S.: Copyright protects creative human selection/arrangement, not raw data or fully autonomous AI output.
Europe: Sui generis database rights protect substantial investment, even with algorithmic assistance.
Global Implication: Organizations must document human contribution, investment, and curation processes to secure ownership rights.
RELATED Blog
comments