IP Challenges In AI-Crafted Predictive Models For Dengue Outbreaks.
1. Understanding AI Predictive Models for Dengue
AI-crafted predictive models for dengue outbreaks analyze epidemiological data, climatic conditions, population mobility, and historical outbreak data to forecast disease trends. These systems are valuable for:
Early warning and public health planning
Resource allocation (hospital beds, vaccines, mosquito control)
Policy decisions and research
IP Challenges arise because such AI models involve:
Data Ownership – Health data is often proprietary or sensitive; access and reuse are legally regulated.
Algorithmic IP – Proprietary AI models may be patentable, but only if they are novel and non-obvious.
Model Output Ownership – Questions about whether predictions themselves can be protected.
Trade Secrets – Epidemiological AI methods may be protected as confidential business information.
2. Key IP Issues in AI-Crafted Predictive Models
Copyright and Data Rights
Public health datasets may be compiled from multiple sources. Copyright may protect the curation or annotation of datasets.
Example: Combining weather, hospital, and mosquito population data into a comprehensive dataset could create copyrightable works.
Patents
AI models predicting dengue outbreaks could be patented if they demonstrate a novel algorithm or method.
Challenges: Many predictive methods are considered abstract ideas, making patent protection difficult (see Alice Corp. case below).
Trade Secrets
Proprietary AI models or predictive methodologies may be protected as trade secrets, especially if kept confidential.
Authorship & Ownership of Outputs
Predictions generated by AI are generally not copyrightable unless there is substantial human authorship.
3. Relevant Case Laws
Here are six detailed cases illustrating IP challenges in AI, data analytics, and predictive modeling:
Case 1: Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991)
Facts: Feist copied data (names and numbers) from Rural’s phone directory. Rural claimed copyright infringement.
Holding: Facts alone cannot be copyrighted, only the creative selection or arrangement.
Relevance:
Epidemiological datasets used for dengue prediction may contain raw facts (infection counts, weather data) that cannot be copyrighted.
AI developers can use factual public data without infringing copyright.
Case 2: Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014)
Facts: Alice Corp. claimed patents for a computer-implemented method for mitigating settlement risk.
Holding: Abstract ideas implemented on a computer are not patentable without an inventive concept.
Relevance:
AI predictive models for dengue may be seen as abstract statistical methods.
Patents are only viable if the model includes a technical improvement or novel algorithm, not just predictive correlation.
Case 3: Authors Guild v. Google, Inc., 804 F.3d 202 (2nd Cir. 2015)
Facts: Google scanned books to create a searchable database; authors sued.
Holding: The use was transformative and fair use.
Relevance:
Aggregating public health data for predictive modeling may qualify as transformative use, especially for research and public health purposes.
This provides legal cover for AI models analyzing dengue outbreak trends using public datasets.
Case 4: SAS Institute Inc. v. World Programming Ltd., [2013] EWCA Civ 1482 (UK)
Facts: World Programming developed software compatible with SAS Institute’s programs without copying the code.
Holding: Functional aspects of software are not copyrightable, only the literal code.
Relevance:
Dengue AI models’ functionality—predicting outbreaks—is generally not protected by copyright, but the code and unique architecture may be.
Reverse-engineering the algorithm is legally permissible if code is not copied.
Case 5: Thaler v. US Copyright Office (2023)
Facts: Copyright claim for AI-generated works (“DABUS”).
Holding: AI-generated works without human authorship cannot be copyrighted.
Relevance:
Predictive outputs of dengue AI models are not eligible for copyright.
Ownership rests on the human creator of the model or dataset.
Case 6: Myriad Genetics, Inc. v. Association for Molecular Pathology, 569 U.S. 576 (2013)
Facts: Myriad patented isolated BRCA1/BRCA2 genes associated with cancer. Court invalidated patents on naturally occurring genes.
Holding: Naturally occurring phenomena cannot be patented; synthetic or modified inventions can.
Relevance:
Epidemiological patterns (e.g., mosquito breeding cycles, climate correlations) are natural phenomena and cannot be patented.
AI models that merely predict based on these natural patterns cannot claim patent solely on discovery of correlations.
Case 7: Oracle v. Google, 872 F.3d 117 (Fed. Cir. 2017)
Facts: Google used Java APIs to develop Android; Oracle claimed copyright infringement.
Holding: Certain functional elements may be copyrightable; fair use applied.
Relevance:
AI models for dengue outbreaks may replicate functional aspects of existing analytics software without infringing IP.
Distinguishes between functional methods and creative implementations.
4. Key Takeaways
Data Ownership: Public epidemiological data can be used freely; curated datasets may have copyright protection.
Algorithm Protection: Patents are limited to novel, technical innovations; abstract statistical or predictive methods may not qualify.
AI Outputs: Predictions themselves cannot be copyrighted; human contribution is essential.
Trade Secrets: Proprietary model architecture and methods can be protected if kept confidential.
Fair Use & Transformative Analysis: AI use for research, public health, or analytics is often legally defensible.
Conclusion
IP challenges in AI predictive models for dengue are similar to those in judicial analytics:
Facts & natural phenomena cannot be patented or copyrighted.
Algorithms and code may be protected, but outputs typically cannot.
Data curation, annotation, and human contribution are key for IP protection.
Courts favor transformative, research-oriented AI use for public health purposes.
RELATED Blog
comments