Part II: AI Patent Challenges in the Insurance and Workers’ Compensation Industry 

The U.S. Patent and Trademark Office (USPTO) released a guidance update on patent subject matter eligibility, especially focused on the consideration of AI. The update, effective from July 17 through Sept. 16 allows public comments to assist in the shaping of patent examiners’ considerations and applicants’ understanding of how to assess AI-related inventions under U.S. patent law or how it may be considered in the future. It builds on previous guidance, offering better clarity and consistency for evaluating how AI inventions may qualify as patentable concepts. 

The USPTO provided this update to offer clearer guidance on evaluating the patent eligibility of inventions, particularly in emerging fields like artificial intelligence (AI), under U.S. patent law (35 U.S.C. § 101). As AI technology advances, determining what qualifies for patent protection has become more complex and confusing, especially when abstract ideas or algorithms are involved, as discussed in Part I of this article series. This signals the broader USPTO initiative to promote guidance on AI development and appropriate incorporation into the inventive process, ensuring that inventors and attorneys have clear and consistent guidelines to follow while still encouraging innovation in critical technological areas. 

A Rapid Transition to New AI Insight and Applications 

The growing application of AI in the insurance and workers' compensation industries is unavoidable, if not already present in practice. AI provides the ability to perform complex data analysis and massive process automation while creating new permutations of understanding by data synthesis (generative AI). The insurance and workers’ compensation industries now face new opportunities to streamline operations and offer new innovative solutions. With the employment of high-accuracy databases (HADs) and validated data sets, the ability to generate higher quality outputs and more accurate data models is now possible. This explosion of computing power coupled with trainable learning will likely unfold the next “technological hook,” bringing the insurance and workers’ compensation industries to a new chapter. The coupling of these two powerful components will lead to many inventions in data text understanding, leading us into areas never before imagined, such as injury impact, secondary data signaling, and ultimately the new frontier of emotional analysis and contextual understanding, which may serve to adjust and advise claim decisions and resources.  

Affective Text Data Analysis Applications 
 
One example of this innovation is the “affective analysis” of report text, where AI algorithms assess the tone, emotional content, and context of reports related to workers' compensation claims. This type of analysis may help insurers more accurately assess the severity of an injury or the impact of a disease, potentially leading to more precise impairment ratings, and better supportive decision-making throughout the claims process. 

In addition to affective analysis (sentiment analysis, emotional recognition, affective computing), AI-powered textual analysis using large, trained datasets will enhance the insurance industry’s ability to predict outcomes and standardize processes across multiple areas of practice. By utilizing large-language models, AI can recognize patterns in claims and medical reports, allowing stakeholders to understand complex concepts, such as injury severity and clinical disease expression, more effectively and quickly. This advanced understanding offers new possibilities for insurance and medical professionals to make more accurate comparisons between cases. Claim aggregations and data cohort organization now usher in a new level of standardization and objectivity in workers' compensation claims. 

Vector Data Analysis Applications 

One of the exciting developments in AI is its capacity to interpret vector values, a key element in natural language processing for understanding signal strength. By analyzing these values, AI can cite subtle nuances of injury descriptions and impact vis-à-vis activities of daily living, indirect pain levels, social impact, and disease progression expression. This capability represents a new resource tool in the insurance and workers' compensation industries, where understanding the degree of injury impact or disease manifestation can significantly influence the outcome of claims, financial reserves, resource utilization, and settlement negotiations. 

AI Patentability Challenges 

However, while AI presents numerous opportunities for innovation, the process of securing patents for AI-driven technologies is fraught with challenges. One major issue is determining whether AI innovations meet the criteria for patentability, including novelty, non-obviousness, and utility. For example, if an AI algorithm improves the accuracy of impairment ratings or automates complex claims processing tasks, it may be eligible for patent protection. However, the novelty requirement means that the invention must differ significantly from existing technologies, and determining this difference in the case of AI can be complicated. Is it acting as a simple computer software model, or is it doing something more? 

AI Non-Obviousness 

“Non-obviousness,” as discussed in Part I, is another critical standard that AI-related inventions must meet to qualify for patents. AI innovations cannot be obvious extensions of current technologies; they must introduce new approaches or algorithms that would not be apparent to experts in the field and carry out a defined action or deliverable. In the case of workers' compensation, for instance, an AI system that automates claims processing might seem like an obvious application of existing technology. But if it uses a unique method, such as employing a novel machine learning model that interprets medical data more accurately than traditional methods, and solves a problem using a novel approach, it could pass the non-obviousness test. 

AI Utility 

The third major criterion for patentability is “utility,” meaning the AI innovation must provide a clear and practical benefit to the industry. In the insurance and workers’ compensation sector, AI can improve decision-making in claims management, predict medical outcomes more accurately, and even assist in pricing and policy adjustments based on real-time data or anticipated data. An AI powered tool that predicts when a workers' compensation claimant will reach Maximum Medical Improvement (MMI) is one such valuable example, offering utility by helping doctors, adjusters, and nurse case managers not only streamline their work but anticipate next steps to ensure claims are processed in a timely and efficient manner. This type of practical innovation aligns well with the utility requirement for patent claims and protection. 

AI and the Alice Decision 

Despite these opportunities, AI patents will likely face significant legal challenges, particularly in relation to overcoming the general concept of being an abstract idea. The U.S. Supreme Court’s decision in the 2014 case Alice Corp. v. CLS Bank International established that abstract ideas, such as “algorithms” or “basic economic principles,” are not patentable unless applied in a novel and specific way. This ruling has major implications for AI in the insurance and workers’ compensation industry, where many innovations involve mathematical algorithms or process automation that may be considered abstract ideas. For an AI algorithm to be patentable, it must solve a specific technical problem in a new way rather than merely automating a known process. 

For example, if an AI tool automates a common task like claims processing, it may be rejected as an abstract idea. However, if the AI tool incorporates a unique algorithm that analyzes claims data in a new way, it may qualify for patent protection. For example, an algorithm might be designed to predict injury severity or disease progression based on real-time data and multiple variables. It might be coupled with specific diagnostic recommendations and treatment interventions with recovery loop analysis, with specific output limits and data ranges of the intervention by way of hardware components comprising a register and a microprocessor. Add in a plurality of synaptic circuits, and together all of these components form an AI invention.  

The AI tool described passes Step 2A of the USPTO subject matter eligibility test under 35 U.S.C. § 101 because it integrates a novel algorithm into an actual practical application. While algorithms can be considered abstract ideas, this invention applies the algorithm to real-time data analysis, specifically predicting injury severity and disease progression, and providing diagnostic recommendations and treatment interventions. Additionally, the inclusion of hardware components, such as a microprocessor, register, and synaptic type circuits, links the abstract idea to a tangible and specific technological solution, making it more than a mere abstract concept. This integration into a concrete system demonstrates that the invention is directed to a practical application, moving beyond a judicial exception. 

This distinction between abstract ideas and patentable inventions is particularly important for AI innovations in the insurance and workers' compensation fields, where many processes involve data interpretation and decision-making. The defining line of the Alice decision challenge remains one of the major litmus tests of patentability. 

AI and Algorithms 

Another major issue that is still being settled is around the patentability of algorithms. While algorithms are central to AI technologies, courts have generally ruled that mathematical algorithms themselves are not patentable unless applied to solve a specific, real-world problem. For instance, an algorithm that simply calculates impairment ratings based on workers' compensation data may not be patentable. But if that algorithm applies novel methods, such as incorporating data from multiple sources to provide a more accurate rating or integrating AI-based predictive analytics to streamline the process while running simultaneous multiple streams of administrative and legal data gates, a process known as “variable thread analytic computation” (VTAC), it could meet the standards for patentability. 

AI and Attorneys 

Given these legal complexities, companies looking to innovate with AI in the insurance and workers' compensation industries must carefully consider their patent strategies earlier than ever. It is essential to ensure that any AI-driven innovation is not only novel and useful but also applied in a way that goes beyond simply automating existing processes. Working with experienced patent attorneys who are well familiar and specialize in AI and intellectual property law is critical to navigating these challenges and securing patents that offer true protection and a solid competitive advantage. 

Innovation Through Patent Law Navigation 

In conclusion, the integration of AI into the insurance and workers' compensation sectors offers vast potential for innovation, but also presents significant challenges and complications in terms of its description of context and patentability. As AI continues to evolve and play a more prominent role in areas such as insurance, workers’ compensation, impairment rating claims processing, and injury outcome prediction, it is crucial for administrative and medical professionals in these fields to understand the legal landscape and limitations surrounding AI patents. By carefully considering the novelty, non-obviousness, and utility of AI innovations, and working with specialized legal professionals, companies can secure valuable patent protections and maintain a competitive edge as innovators and leaders in the rapidly changing world of insurance and workers' compensation. 

The path to obtaining AI patents may not be straightforward, but with a basic understanding of the strategic approach, stakeholders can navigate these complexities and successfully protect inventions and intellectual property. In the future, as AI becomes even more deeply integrated into the processes of the insurance and workers' compensation industries, the importance of robust patent strategies will only grow in value. Understanding the intersection of AI, intellectual property law, practical and specific applications of data analysis, and the unique challenges of patenting in these industries is critical for staying ahead in this innovative future. 

About the Authors 

John Alchemy, M.D., QME, DABFP, has 30 years of clinical and legal experience in California Workers' Compensation. He is the founder of Impairment Ratings Specialists, A Medical Corporation, and Alchemy Logic Systems, operating as RateFast, an industry-leading software and data analytics platform for workers' compensation. Dr. Alchemy holds eight USPTO patents in the fields of insurance, impairment rating, and database validation and design, with additional patents pending. 

Avery Steffen, J.D., graduated from UCLA Law School in 2023 and is actively practicing in Sacramento, Calif. He has experience in the California workers' compensation administrative system. Mr. Steffen has authored educational articles for the California Medical Unit's continuing education program with the retired judge Hon. Steve Siemers, focusing on impairment rating for qualified medical evaluators. 

The information presented in this article is intended for general informational purposes only and should not be interpreted as legal advice. Laws vary by jurisdiction and may change over time. Readers are encouraged to seek professional legal counsel for specific guidance related to their individual circumstances. This article does not create an attorney-client relationship.

Share This Article: