The U.S. Patent and Trademark Office (USPTO) recently issued new guidance regarding patent-eligible subject matter for artificial intelligence (AI). This new guidance is intended to assist patent examiners and stakeholders in evaluating AI-related patent claims but can also be invaluable for those seeking to protect their AI innovations.
Understanding patent eligibility
At the core of this guidance is the requirement that any inventive concept must fit within one of four statutory categories: process, machine, manufacture, and composition of matter. Additionally, the inventive concept should avoid falling under judicial exceptions, such as abstract ideas, laws of nature, and natural phenomena. The guidance emphasizes two primary considerations under 35 U.S.C § 101:
- Is the claimed invention an abstract idea?
- Does the claim integrate the judicial exception into a practical application?
Examples from the guidance
To illustrate these principles, the USPTO provided three notable examples of AI-related inventions that were deemed patent eligible:
- Artificial neural networks – In this case, the claims pertained to a physical device having specific hardware components and a specific manner of training neural networks. This was considered an improvement to other technology or technical field.
- AI-based speech analysis – The second example disclosed AI-based methods of analyzing speech signals and separating desired speech from extraneous or background speech. The first claim determined to be patent-eligible subject matter had specific structure and training of the machine learning model. Thus, the claim was found to integrate the abstract idea into a practical application. The second claim found to be patent-eligible subject matter focused on the trained model rather than the training process, demonstrating an improvement in speech signal processing technology.
- Fibrosis treatment using AI – The third example disclosed AI-assisted personalized medical treatments. In this case, the first claim was determined not to be patent eligible because the claimed subject matter used a generic machine learning model using a generic computer component. In contrast, the claim that was determined to have patent eligible subject matter added details about the treatment utilizing a specific compound. In other words, the claimed subject matter integrated the abstract idea into a practical application having meaningful limits.
Key takeaways for pursuing AI patent protection
While pursuing AI-related subject matter can be challenging, it is certainly feasible. When considering whether to pursue patent protection for your AI innovations, keep the following points in mind:
- Technological improvement: Does the inventive concept improve the functioning of a computer or other technology in the technical field?
- Problem solving: Does the inventive concept provide a solution to a problem, rather than merely claiming the idea of a solution?
Aligning the patent claims with these criteria in mind, you can strengthen your position in the pursuit of a patent for AI-related innovations.
