Ownership Of Predictive Algorithms For Sustainable Urban Infrastructure Development.
I. Understanding Predictive Algorithms in Sustainable Urban Infrastructure
Predictive algorithms for sustainable urban infrastructure are AI-based systems that:
Forecast traffic flows, energy consumption, and water usage.
Optimize waste management, public transport, and building energy efficiency.
Analyze sensor, IoT, and environmental data to inform urban planning decisions.
Enable smart city solutions with sustainable outcomes.
Key Ownership Concerns
Intellectual Property Ownership: Who owns the algorithm—developer, city, contractor, or AI system itself?
Copyright vs. Patent Protection: Can predictive algorithms be patented, copyrighted, or considered trade secrets?
Collaborative Development: Often algorithms are developed jointly by public and private entities.
Data Ownership: Algorithms rely on data; questions arise whether data contributors hold any rights.
International Deployment: Cross-border use raises additional legal complexities.
II. Legal Themes Relevant to Ownership
Patent Protection of Algorithms – Must demonstrate technical innovation, novelty, and utility.
Copyright & Software Protection – Code may be copyrightable if human-authored.
Trade Secrets – Ownership can be protected as confidential business information.
Collaborative Inventions – Determining inventorship and joint ownership.
Data & AI Ownership – Ownership of derivative insights versus raw data.
III. Key Case Laws (Detailed Analysis)
1. Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014)
Facts: Alice Corp. sought a patent on a computer-implemented financial method.
Held: Abstract ideas implemented on computers are not patentable unless the claim includes an inventive concept.
Relevance:
Predictive algorithms for urban planning may be deemed abstract ideas.
Patent protection requires demonstrating technical implementation beyond standard computation, e.g., integrating IoT sensors, optimization engines, or real-time prediction pipelines.
2. Diamond v. Chakrabarty, 447 U.S. 303 (1980)
Facts: The Supreme Court considered patenting genetically engineered bacteria.
Held: Human-engineered inventions that are novel and non-naturally occurring are patentable.
Relevance:
Predictive algorithms with novel architectures, modeling techniques, or AI integration may qualify for patents if they solve technical problems in urban infrastructure sustainably.
3. Thaler v. U.S. Copyright Office, 2023 (Stephen Thaler AI Case)
Facts: Thaler claimed copyright for works autonomously generated by AI.
Held: Human authorship is required; AI cannot hold copyright independently.
Relevance:
Ownership of predictive algorithms requires human inventorship or authorship.
Even if AI contributes to algorithm design, ownership rests with humans/entities guiding development.
4. Gottschalk v. Benson, 409 U.S. 63 (1972)
Facts: Court addressed a patent for a method converting decimal numbers to binary.
Held: Mathematical algorithms alone are not patentable.
Relevance:
Purely predictive formulas or models for urban infrastructure cannot be patented.
Must show practical application, e.g., system controlling smart traffic lights or energy grids.
5. Google LLC v. Oracle America, Inc., 593 U.S. ___ (2021)
Facts: Google used Java APIs in Android; Oracle claimed copyright infringement.
Held: The Supreme Court ruled that Google’s use was fair use, emphasizing functional nature and interoperability.
Relevance:
Predictive algorithm ownership can be affected when standardized APIs, shared datasets, or modular frameworks are used.
Licensing and interoperability agreements are critical for urban infrastructure AI deployment.
6. Schrems II, CJEU, 2020
Facts: EU-US Privacy Shield invalidated due to insufficient protection of personal data.
Relevance:
Predictive algorithms rely on sensitive urban data (traffic patterns, energy usage, public mobility).
Ownership is intertwined with data protection compliance; unauthorized use of data could challenge algorithm ownership or commercial rights.
7. In re Nuijten, 500 F.3d 1346 (Fed. Cir. 2007)
Facts: Patent claimed a transitory signal via a transmission medium.
Held: Claims directed solely to signals are not patentable.
Relevance:
Predictive algorithm outputs (forecasts, signals) cannot be patented as standalone information.
Protection must cover the system, method, or device implementing the algorithm, not just the predictions.
8. Diamond v. Diehr, 450 U.S. 175 (1981)
Facts: A process for curing rubber used a computer algorithm to calculate timing.
Held: Process was patentable because it applied a formula in a practical technological process.
Relevance:
Predictive urban algorithms are patentable if they solve a technical problem in a concrete application, e.g., energy load balancing or traffic flow optimization.
IV. Doctrinal Principles for Ownership of Predictive Algorithms
| Principle | Implication for Urban Infrastructure AI |
|---|---|
| Patent Eligibility | Technical implementation and practical application required (Alice, Diehr) |
| Human Inventorship | AI alone cannot hold patent or copyright (Thaler) |
| Novelty & Non-Obviousness | Unique algorithm architectures may be patented (Chakrabarty) |
| Data Compliance | Ownership depends on lawful access to sensitive urban data (Schrems II) |
| Trade Secrets & Licensing | Proprietary datasets and modular frameworks require contracts (Google v. Oracle) |
| Abstract vs. Applied Algorithms | Forecasting methods must be embodied in systems for patent protection (Gottschalk, Nuijten) |
V. Key Takeaways
Predictive algorithms can be patented if they involve human ingenuity, technical implementation, and solve practical urban problems.
AI alone cannot be recognized as an inventor; ownership rests with humans or organizations.
Copyright may protect code, but not abstract prediction formulas.
Trade secrets, licensing, and interoperability are crucial for collaborative urban infrastructure projects.
Data privacy compliance affects ownership rights when algorithms rely on personal or sensitive urban datasets.
RELATED Blog
comments