Ipr In Cognitive Computing Applications.
1. Cognitive Computing and IPR: Conceptual Background
What is Cognitive Computing?
Cognitive computing systems are AI-driven technologies that simulate human thought processes. They:
Learn from data (machine learning, deep learning)
Understand natural language (NLP)
Reason and make decisions
Improve autonomously over time
Examples include:
AI medical diagnosis systems
Intelligent chatbots
Predictive legal analytics
Recommendation engines
Autonomous decision-making platforms
Because these systems create, learn, and sometimes invent, they challenge traditional IPR frameworks.
2. Types of IPR Relevant to Cognitive Computing
(a) Copyright
Protects:
Source code
Training data selection and arrangement
Outputs only if human creativity exists
Challenges:
AI-generated works without human authorship
Ownership of machine-generated content
(b) Patents
Protect:
Novel algorithms (if technical effect is shown)
AI-based processes
System architectures
Challenges:
Abstract ideas vs technical inventions
Inventorship when AI creates the invention
(c) Trade Secrets
Protect:
Training datasets
Model weights
Proprietary algorithms
Challenges:
Reverse engineering
Employee mobility
(d) Data Protection and Database Rights
Protect:
Structured datasets used for training
Data curation efforts
Challenges:
Use of public or scraped data
Consent and ownership
(e) Moral and Ethical Dimensions
Accountability for AI decisions
Bias in datasets
Attribution of creative credit
3. Key IPR Issues in Cognitive Computing
Who owns AI-generated output?
Can AI be an inventor or author?
Patentability of algorithms
Use of copyrighted data for training
Liability for infringement by AI systems
4. Case Laws (Explained in Detail)
Case 1: Feist Publications v. Rural Telephone Service (1991)
Issue:
Whether mere compilation of data is protected under copyright.
Facts:
Rural Telephone published a phone directory
Feist copied factual listings into its own directory
Rural claimed copyright infringement
Judgment:
Facts are not copyrightable
Only original selection and arrangement are protected
Relevance to Cognitive Computing:
AI training datasets often consist of factual data
Raw data used to train AI cannot be copyrighted
However, curated datasets with creative arrangement may be protected
Principle Established:
Cognitive computing systems trained on factual databases do not infringe copyright unless creative expression is copied.
Case 2: Eastern Book Company v. D.B. Modak (India)
Issue:
Whether editorial enhancements in legal judgments are copyrightable.
Facts:
Eastern Book Company added headnotes, formatting, and paragraphing to judgments
Competitors copied these features
Judgment:
Copyright exists if minimal creativity is applied
“Sweat of the brow” alone is not enough
Relevance to Cognitive Computing:
Legal AI tools trained on judgments must avoid copying:
Headnotes
Editorial summaries
Enhanced formatting
Principle Established:
AI systems may use public domain judgments but must not replicate proprietary editorial contributions.
Case 3: R.G. Anand v. Deluxe Films (India)
Issue:
Whether copying an idea amounts to copyright infringement.
Facts:
Plaintiff claimed a film copied the theme of his play
Judgment:
Copyright protects expression, not ideas
Similar ideas are allowed if expression differs
Relevance to Cognitive Computing:
AI-generated outputs may resemble existing works
No infringement if:
Only the idea or concept is similar
Expression is independently generated
Principle Established:
Cognitive systems can lawfully generate content inspired by existing works, provided expression is original.
Case 4: Alice Corp. v. CLS Bank International (2014)
Issue:
Are abstract ideas implemented via computers patentable?
Facts:
Alice Corp patented a computerized financial transaction method
Opponent argued it was an abstract idea
Judgment:
Abstract ideas are not patentable
Must demonstrate technical innovation
Relevance to Cognitive Computing:
AI algorithms alone are not patentable
Must show:
Technical advancement
Improved computing efficiency
Hardware interaction
Principle Established:
Cognitive computing inventions must show a technical contribution beyond mathematical logic.
Case 5: DABUS Artificial Intelligence Inventorship Cases
Issue:
Can an AI system be recognized as an inventor?
Facts:
DABUS AI created inventions without human input
Patent applications named AI as inventor
Decisions:
Patent offices rejected applications
Inventor must be a natural person
Relevance to Cognitive Computing:
AI cannot own patents
Ownership lies with:
Developer
Owner
Employer
Principle Established:
Cognitive computing systems cannot be legal inventors under current IPR laws.
Case 6: Google LLC v. Oracle America Inc. (2021)
Issue:
Whether copying API declarations amounts to copyright infringement.
Facts:
Google copied Java API structure
Oracle sued for infringement
Judgment:
Copying APIs was fair use
Encouraged innovation and interoperability
Relevance to Cognitive Computing:
AI developers often rely on APIs
Training or using standardized interfaces may be lawful
Principle Established:
Functional elements used for innovation may fall under fair use in AI development.
Case 7: Cambridge Analytica–Facebook Data Use Controversy (Principle-Based)
Issue:
Unauthorized data use for AI-driven profiling
Key Outcome:
Emphasized consent and data ownership
Highlighted ethical limits of cognitive systems
Relevance:
Training cognitive systems on personal data can violate data protection laws
IPR must coexist with privacy rights
5. Comparative Summary of Legal Principles
| Issue | Legal Position |
|---|---|
| AI as author/inventor | Not recognized |
| Data used for training | Facts allowed, expression restricted |
| Patentability | Requires technical effect |
| AI-generated output | Ownership lies with humans |
| Dataset protection | Depends on creativity and structure |
6. Conclusion
Cognitive computing presents unprecedented challenges to traditional IPR frameworks. Current laws:
Do not recognize AI as a legal person
Focus on human creativity and control
Require balancing innovation with rights protection
Future reforms may need to:
Introduce AI-specific authorship rules
Clarify ownership of machine-generated works
Address ethical and accountability concerns
RELATED Blog
comments