IP Disputes In Poland Involving AI-Personalised Fitness Animations.
1. Polish Supreme Court Case on Originality
(Case No. II CSK 400/16, Supreme Court of Poland, 2017)
Facts
The dispute concerned whether a conceptual system created by combining existing elements could be protected under copyright law. The case did not involve AI, but it established the test for originality, which is crucial for AI-generated works such as fitness animations.
A party claimed copyright protection over a concept built from already existing elements.
Legal Issue
The court examined:
Whether combining existing components can constitute a creative work
Whether originality requires complete novelty or merely a unique combination
Judgment
The Polish Supreme Court ruled that copyright protection exists if a work shows individual creativity, even when it is composed of previously known elements.
The court held that originality exists if the work is statistically unlikely to be reproduced identically by another person.
Legal Principles Established
The judgment clarified that:
Originality may arise from creative arrangement of existing elements.
Protection depends on individual creative input.
Pure ideas are not protected, only their expression.
Relevance to AI-Fitness Animations
This case is important because AI fitness animation systems often combine:
motion-capture data
exercise databases
animation templates
biomechanical models
Under this case law:
The creative arrangement of movements and animations may be protected.
However, mere algorithmic combinations by AI without human creativity may not qualify as copyrightable works.
2. AI-Generated Artwork Copyright Dispute in Europe (Referenced in Polish Legal Commentary)
Facts
A claimant attempted to enforce copyright over an AI-generated image produced using an automated generative system.
The user argued that the prompt they entered into the AI system constituted authorship.
Legal Issue
The court examined:
Whether AI-generated works qualify as “creative works”.
Whether prompt engineering can establish authorship.
Judgment
The court ruled that:
AI cannot be recognized as an author.
The claimant failed to demonstrate sufficient human creative input.
The prompt itself was considered an idea rather than a creative work.
Therefore, the work was not protected by copyright.
Legal Principles
The case established that:
Only natural persons can be authors.
AI-generated content without human creativity lacks copyright protection.
Users must demonstrate decisive creative control.
Relevance to AI Fitness Animation Systems
Fitness platforms using AI to generate:
animated trainers
motion-correcting avatars
personalized exercise visuals
may face the same legal issue:
If the animation is produced entirely by AI, it may not be copyrightable.
Consequences:
Companies cannot enforce exclusive rights over such animations.
Competitors could potentially reuse identical animations.
3. Newag v. Dragon Sector Software Dispute (Warsaw District Court, 2024)
Facts
A major dispute arose between Newag, a train manufacturer, and Dragon Sector, cybersecurity researchers.
Dragon Sector accessed and analyzed proprietary software embedded in Newag trains to identify potential technical restrictions.
Newag sued for:
copyright infringement
unauthorized access to proprietary software.
Legal Issues
The court had to determine:
Whether reverse engineering software constitutes copyright infringement.
Whether analyzing software without modifying it violates IP rights.
Court Proceedings
During trial, Newag acknowledged that Dragon Sector did not actually modify the software, but argued that analyzing it was unauthorized.
The case therefore focused on the legality of software analysis and reverse engineering.
Legal Significance
The case highlights tensions between:
copyright protection of software
legitimate research and analysis.
Relevance to AI Fitness Animation Platforms
AI fitness animation systems rely on complex proprietary software including:
motion-generation engines
pose-estimation algorithms
biomechanical AI models.
If a competitor reverse engineers these systems, similar disputes may arise involving:
copyright infringement
software analysis rights
trade secret protection.
4. Text and Data Mining (TDM) Dispute Involving AI Training Data
Background
AI models require large datasets for training, including:
exercise videos
motion capture recordings
animation libraries
athlete performances.
Under EU Directive 2019/790 (DSM Directive), text and data mining allows AI training in certain circumstances.
However, Polish copyright stakeholders—including collective rights organizations—have challenged unrestricted AI training.
Legal Issue
The dispute concerns whether AI developers can:
train models using copyrighted material
without explicit permission.
Rights holders argue that:
AI training constitutes reproduction of protected works.
licensing is required.
Legal Consequences
If courts rule in favor of rightsholders:
AI developers must obtain licenses for training datasets.
unauthorized training could lead to copyright infringement claims.
Relevance to AI Fitness Animations
Fitness AI models are often trained on:
professional workout videos
sports motion datasets
yoga and training tutorials.
If these materials are copyrighted, training AI systems without authorization could lead to major IP disputes.
5. Authorship and Liability in AI-Generated Works under Polish Copyright Law
Legal Framework
Polish copyright law defines a protected work as:
a manifestation of creative activity of an individual character.
This definition implies that human intellectual effort is required.
AI systems therefore:
cannot hold copyright
cannot be liable for infringement.
Liability Questions
When AI produces infringing content, responsibility may fall on:
the developer of the AI system
the platform operating the system
the user providing prompts or instructions.
Application to Fitness Animation Platforms
Suppose a fitness app generates animations that closely replicate choreography from a copyrighted training video.
Possible legal claims:
copyright infringement
unfair competition
derivative work violation.
Liability may be assigned to:
the company providing the AI tool
the developer who trained the model
the user generating the animation.
Key Legal Issues Specific to AI-Personalised Fitness Animations
1. Ownership of Generated Animations
If the animation is fully AI-generated, it may not receive copyright protection.
2. Copyright in Movement or Exercise Routines
Courts generally protect:
expressive choreography
detailed motion sequences.
But simple exercises may not be protected.
3. Training Data Liability
Using copyrighted workout videos for training AI may create copyright infringement claims.
4. Software Protection
The AI software generating the animations is protected as a computer program.
5. Reverse Engineering Disputes
Competitors studying the system could trigger software copyright litigation.
Conclusion
IP disputes involving AI-personalised fitness animations in Poland revolve around several unresolved legal questions:
Authorship – AI cannot be an author; human creativity is required.
Ownership – purely AI-generated animations may lack copyright protection.
Training data – using copyrighted exercise videos for AI training may infringe copyright.
Software rights – the algorithms generating animations remain protected.
Liability – responsibility for infringement may fall on developers, platforms, or users.
The discussed case laws and disputes show that Polish courts emphasize human creativity and originality, making AI-generated works legally uncertain and likely to remain a major field of intellectual-property litigation.
RELATED Blog
comments