Patent Governance For Neuro-Adaptive Computing Applied To Creative Production.
I. Conceptual Framework: Neuro-Adaptive Computing in Creative Production
Neuro-adaptive computing refers to computing systems inspired by neural networks and adaptive algorithms, capable of learning and evolving over time. When applied to creative production, this includes:
- Generative art, music, and content systems
- AI-assisted design tools (architecture, industrial design)
- Adaptive storytelling or game content generation
- Human-AI collaboration platforms
Patent governance for such technologies is complex because they intersect software, AI, and creativity domains. Key considerations:
- Patent Eligibility: Are these algorithms abstract ideas, or do they produce a technical effect?
- Inventorship: Can AI be considered an inventor?
- Technical Contribution: Does the AI system improve computing machinery or creative workflows materially?
- Scope and Enforcement: Protecting outputs vs. protecting system architecture.
II. Landmark Case Laws (Detailed Explanation)
1. Alice Corp. v. CLS Bank (2014)
Facts
- Patent involved computerized financial transaction systems.
Issue
- Whether software implementing a process on a computer is patentable.
Judgment
- Supreme Court invalidated the patent.
Principle
- Introduced the Alice/Mayo two-step test:
- Determine if the claim is an abstract idea
- Determine if there is an inventive concept beyond the abstract idea
Relevance to Neuro-Adaptive Creative Systems
- Algorithms generating music or visual content may be considered abstract ideas unless applied to technical devices or processes.
- Example: Simply outputting images or music based on AI training data is likely not patentable.
2. Mayo Collaborative Services v. Prometheus (2012)
Facts
- Patent claimed correlations between drug dosage and patient outcomes.
Judgment
- Invalidated because it claimed a natural law.
Principle
- Natural laws are not patentable, unless applied in a novel and practical way.
Relevance
- Creative AI systems that rely solely on patterns in data may be considered natural correlations.
- Patentable claims must demonstrate novel technical implementation, like improving computing efficiency in generating creative outputs.
3. Thaler v. Comptroller-General (DABUS Case, UK / US 2021-2022)
Facts
- AI system “DABUS” listed as inventor for a beverage container and flashing light patent.
Judgment
- Courts (UK, US, EU) ruled AI cannot be an inventor.
Principle
- Only human beings can be legally recognized as inventors.
Relevance
- Neuro-adaptive creative AI may generate original work, but the human supervising or programming the system must be listed as the inventor.
4. Electric Power Group v. Alstom (2016)
Facts
- Patent for monitoring electric power grids using data analytics.
Judgment
- Invalidated as abstract.
Reasoning
- The system only collected, analyzed, and displayed information, without technological innovation.
Relevance
- AI creative production systems that simply generate outputs from datasets may be rejected unless they improve hardware, reduce computational load, or optimize processing in a novel way.
5. Flook v. Parker (1978)
Facts
- Patent claimed a method for updating alarm limits in a chemical process using a formula.
Judgment
- Supreme Court invalidated it.
Principle
- Application of a formula is not enough; there must be a novel, technical implementation.
Relevance
- In creative AI, merely using neural networks or adaptive algorithms is not enough.
- Patentable claims must show specific application or improvement in a process or system, e.g., faster rendering, adaptive design pipelines.
6. Association for Molecular Pathology v. Myriad Genetics (2013)
Facts
- Patents on isolated human DNA sequences.
Judgment
- Natural DNA = not patentable, but synthetic DNA (cDNA) = patentable.
Principle
- Human modification or engineering of a natural phenomenon is patentable.
Relevance
- AI training datasets (raw data) are analogous to “natural phenomena”.
- Novel neuro-adaptive architecture or engineered creative algorithms are patentable.
7. Thales v. Iancu (DABUS US Fed. Circuit, 2022)
Facts
- AI-generated inventions submitted for patents.
Judgment
- AI alone cannot hold patent rights; humans must be named.
Key Point
- Reinforces the human inventorship rule, even in highly creative AI systems.
III. Key Governance Principles
- Abstract Idea Limitation
- AI algorithms producing creative content may be abstract unless they improve computing or processes.
- Inventive Concept Requirement
- Must show technical effect, e.g., hardware acceleration, novel neural architecture.
- Human Inventorship Rule
- AI cannot be inventor; humans controlling/training the AI must be listed.
- Technical Implementation over Output
- Focus on how creative output is generated, not just the content.
- Data as Input vs. Engineered System
- Raw datasets are not patentable; system improvements are.
IV. Application to Neuro-Adaptive Creative Production
Patentable Example
✔ AI-assisted design tool that:
- uses neuro-adaptive computing to optimize 3D printing paths
- reduces material waste and printing time
- adaptively modifies models in real-time
Non-Patentable Example
✘ AI algorithm that:
- generates art images based on existing paintings
- without improving computational efficiency or process
V. Emerging Issues
- Ownership of AI-generated creative outputs
- Scope of patent claims for systems that combine AI + human creativity
- Cross-jurisdiction differences (US vs EU vs UK) on AI inventorship
VI. Conclusion
Patent governance for neuro-adaptive computing in creative production depends on:
- Alice/Mayo test (abstract idea + inventive concept)
- Technical implementation focus
- Mandatory human inventorship
- Exclusion of natural or unmodified datasets
Innovation must demonstrate a technical or procedural improvement, not just generative creativity.
RELATED Blog
comments