Dr. Terri Shieh-Newton, an intellectual property attorney and immunologist at Mintz, discusses AI-driven healthcare solutions’ ownership rights. She highlights the importance of probing data methodologies and classification for robust patents. The challenge of establishing AI-related intellectual property ownership is addressed, emphasizing inventor identification and data curation. Collaborations and data exchanges are recognized as areas prone to ownership complexities, necessitating meticulous agreements to ensure progress while averting disputes.
The field of healthcare has seen an increase in the use of artificial intelligence and large language models in the realm of AI-driven solutions. This trend raises questions about who owns the intellectual property associated with innovations born from these emerging technologies.
Dr. Terri Shieh-Newton, a distinguished intellectual property attorney, accomplished immunologist, and integral member of the law firm Mintz, has taken center stage in addressing these legal conundrums. Notably, she spearheads the Life Sciences Artificial Intelligence group at Mintz, a specialized team comprising experts such as microbiologists, physicists, immunologists, chemists, electrical engineers, and computer scientists. These professionals convene monthly to delve into the intricacies of patents and concepts concerning novel AI models.
In an insightful conversation with MobiHealthNews, Dr. Shieh-Newton elucidates the legal aspects of establishing intellectual property rights within the context of AI deployment and guides companies navigating ownership intricacies.
MobiHealthNews: In the context of healthcare, particularly in ensuring unbiased data use alongside AI, what counsel do you offer clients?
Dr. Terri Shieh-Newton: As an intellectual property attorney, my primary focus isn’t database design; that’s the domain of data scientists. However, collaborating with them allows me to ask inquisitive questions about data assembly, training methodologies, and exclusion criteria. My role involves probing into the specifics of data classification, as it significantly influences patent strength. In light of the Amgen v. Sanofi ruling, it’s essential to consider the scope of claims against available data. With machine learning and extensive datasets, the possibility of refining patent descriptions arises. Addressing questions about dataset origins, potential biases, and exclusions becomes paramount for bolstering patent breadth.
MHN: Can you elaborate on how companies can establish intellectual property ownership as AI solutions evolve?
Shieh-Newton: The landscape remains largely uncharted, with no established laws. The crux lies in identifying the individual behind the algorithm’s creation and data training. The model’s nature, whether supervised learning or otherwise, influences curation considerations. Effective data curation and the deliberate separation of modules can determine inventorship, as AI cannot be deemed an inventor. This underscores the importance of robust employment and inventorship assignment agreements, given the evolving legal landscape. Unlike certain jurisdictions, the automatic ownership of data isn’t a given.
MHN: For companies without such contracts, ownership disputes might surface.
Shieh-Newton: Contracts addressing employment typically entail standardized clauses. Murkier waters surround data ownership. Collaborations and data exchanges are rife, making data quantity integral to training. However, pinpointing data origins becomes intricate when institutions and companies collaborate. Tracking data sources, ensuring HIPAA compliance, and clarifying data sharing permissions pose challenges. This scenario transcends patents, posing complex concerns about data provenance. Advanced planning, comprehensive agreements, and mutual intentions can avert conflicts and pave the way for progress and shared advancements in medicine.
In the grand endeavor to foster medical breakthroughs, aligning intentions and fostering collaboration necessitates a diligent approach, fortified by proactive agreements and meticulous planning.
