Copyright Implications For AI-Generated Architectural Film Content.
1. AI-Generated Content and Copyright Basics
Copyright law traditionally protects original works of authorship fixed in a tangible medium. In architecture and film:
Architectural designs themselves can be protected as architectural works.
Films documenting architecture are protected as cinematographic works.
Copyright protection requires human authorship. AI complicates this because it can generate images, animations, or films based on prompts without direct human authorship.
Key legal questions in AI-generated architectural films:
Who owns the copyright—the AI user, the AI developer, or no one?
Can architectural works used as input in AI infringe original copyrights?
How does fair use apply in documentary or artistic contexts?
Are derivative works protected if based on copyrighted architecture?
2. Case Law Examples
Case 1: Naruto v. Slater (2018, U.S.) – “Monkey Selfie”
Facts: A monkey took a selfie using a photographer’s camera. The photographer tried to claim copyright.
Relevance: U.S. courts held that copyright protection requires human authorship. Animals or AI cannot own copyright.
Implication for AI films: If AI generates architectural content entirely autonomously, it may not be copyrightable, though human selection or editing may qualify.
Case 2: Feist Publications, Inc. v. Rural Telephone Service Co. (1991, U.S.)
Facts: Rural Telephone Service compiled a phone directory. Feist copied portions.
Holding: Facts are not copyrightable; copyright requires minimal originality.
Implication: AI-generated architectural content must involve some human creative input to qualify. Purely AI-generated renderings may be deemed uncopyrightable facts or data.
Case 3: Warhol Foundation v. Goldsmith (2021, U.S.)
Facts: Andy Warhol used a copyrighted photograph to create silkscreen images.
Holding: Courts focused on transformative use and substantial similarity.
Implication: Using copyrighted architectural images as prompts for AI films could infringe if the AI output is not transformative enough. Merely stylizing or animating a building may not be safe.
Case 4: Thaler v. Comptroller-General of Patents, Designs, and Trade Marks (UK, 2022)
Facts: Stephen Thaler argued that an AI (DABUS) should be recognized as an inventor.
Holding: Courts rejected non-human inventors; IP law requires human authorship.
Implication: Similar reasoning applies to architectural AI films. Human curation, direction, or editing may be necessary for copyright.
Case 5: Castle Rock Entertainment v. Carol Publishing Group (1998, U.S.)
Facts: A trivia book copied scenes from “Seinfeld”.
Holding: Courts considered the substantial similarity test in derivative works.
Implication: If an AI architectural film copies copyrighted photos or designs too closely, it may be a derivative work, risking infringement.
Case 6: Infopaq International A/S v. Danske Dagblades Forening (CJEU, 2009)
Facts: Court considered the copyrightability of brief extracts of text.
Holding: Even short extracts can infringe if they are original expression.
Implication: Portions of architectural designs or blueprints used in AI films, even small details, could trigger copyright issues if recognizable.
Case 7: Andy Warhol Foundation v. Goldsmith (2023, expanded context)
This reinforces that AI outputs using copyrighted images are scrutinized under substantial similarity and transformative use. Courts are cautious about creative contributions.
3. Practical Takeaways for AI-Generated Architectural Films
Human authorship matters:
Only films with substantial human creative input (storyboarding, editing, selection) are likely to be copyrightable.
Avoid direct copying:
Using copyrighted building photos or designs as AI input can create derivative works that infringe.
Transformative use defense:
If the AI-generated content transforms the original architecture substantially (animation, reinterpretation, stylization), it may be safer, but this is legally uncertain.
License or public domain:
Use architecture images from licensed sources, public domain, or your own designs to avoid risk.
Document human involvement:
Courts may consider the extent of human decision-making in AI films—this can strengthen copyright claims.
4. Summary Table of Case Implications
| Case | Principle | Implication for AI Architectural Film |
|---|---|---|
| Naruto v. Slater | Only humans can hold copyright | Purely AI-generated film may not be copyrighted |
| Feist v. Rural | Minimal originality required | Human creativity needed for copyright |
| Warhol v. Goldsmith | Transformative use vs. derivative | AI using copyrighted architecture may infringe |
| Thaler v. UK IPO | Non-human cannot be inventor | AI alone cannot claim copyright |
| Castle Rock v. Carol | Substantial similarity | AI copying architectural works may infringe |
| Infopaq | Even small extracts can infringe | AI using detailed design elements risky |
Conclusion:
AI-generated architectural films occupy a legal gray zone. Copyright protection is strongest when humans provide meaningful creative input, while purely autonomous AI output is unlikely to be protected. Using copyrighted architectural works without modification risks infringement, especially under derivative work doctrines. Courts continue to clarify these principles through cases involving AI, transformative art, and derivative works.
RELATED Blog
comments