Machine learning is increasingly becoming a critical component of many technologies, and its patentability has been a topic of debate in the United States. The Federal Circuit recently addressed this issue in the case of Recentive Analytics, Inc. v. Fox Corp., et al., 2023-2437, ruling on the validity of certain machine-learning patents under Section 101. The patents in question used machine learning to optimize live-event schedules and network maps, which are determined by a broadcaster’s channels within certain geographic markets at particular times. The patents claimed that these optimizations were achieved through the application of established machine learning methods to a new data environment, raising questions about their patent eligibility under Section 101. The court used the two-step Alice framework, which is commonly applied in software patent cases. The first step, Alice Step 1, was directed to determining whether the claims were directed to abstract subject matter. The court held that the disputed claims were indeed directed to abstract subject matter, as they relied on the use of generic machine learning technology in carrying out the claimed methods. The second step, Alice Step 2, was concerned with determining whether the claims contained an inventive concept. The court found that the claims did not meet this threshold, as they did not delineate steps through which the machine learning technology achieved an improvement. The Federal Circuit’s decision highlights the importance of demonstrating a technological improvement in machine learning patents. The court emphasized that machine learning is a burgeoning field and may lead to patent-eligible improvements in technology. However, the court also noted that patents claiming only the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models, are not patent eligible. The decision has significant implications for the patentability of machine learning patents. It underscores the need for patent applicants to demonstrate a technological improvement in their machine learning claims, rather than simply applying established machine learning methods to new data environments. For example, consider the case of a patent applicant who claims to have developed a machine learning algorithm that optimizes event schedules based on a user’s target features. While this may be a significant improvement over existing algorithms, the court may still find that the patent is not eligible for patent protection. On the other hand, consider a patent applicant who claims to have developed a machine learning algorithm that not only optimizes event schedules but also improves the accuracy of the predictions. This would likely be considered a technological improvement and would be eligible for patent protection. In conclusion, the Federal Circuit’s decision in Recentive Analytics v. Fox Corp. provides valuable guidance on the patentability of machine learning patents. It highlights the importance of demonstrating a technological improvement in machine learning claims, rather than simply applying established machine learning methods to new data environments.

“Today, we hold only that patents that do no more than claim the application of generic machine learning to new data environments, without disclosing improvements to the machine learning models to be applied, are patent ineligible.”

The decision also highlights the importance of considering the context in which machine learning is being applied. The court observed that the patents in question relied on the use of generic machine learning technology in carrying out the claimed methods, and that the claimed machine learning technology was conventional. The court also rejected Recentive’s argument that the claimed methods’ application of machine learning to a new field of use conferred eligibility. The court recited its black-letter rule that “an abstract idea does not become nonabstract by limiting the invention to a particular field of use or technological environment.”

The decision has significant implications for the patentability of machine learning patents, particularly for those that claim to use machine learning in a novel way. It underscores the need for patent applicants to carefully consider the context in which they are applying machine learning, and to ensure that their claims meet the requirements of patent eligibility. For example, consider the case of a patent applicant who claims to have developed a machine learning algorithm that uses data from social media to predict user behavior. While this may be a novel application of machine learning, the court may still find that the patent is not eligible for patent protection if the application of machine learning does not constitute a technological improvement. In contrast, consider a patent applicant who claims to have developed a machine learning algorithm that not only uses data from social media but also improves the accuracy of the predictions by incorporating additional data sources. This would likely be considered a technological improvement and would be eligible for patent protection. Ultimately, the Federal Circuit’s decision in Recentive Analytics v. Fox Corp. provides valuable guidance on the patentability of machine learning patents. It highlights the importance of demonstrating a technological improvement in machine learning claims, and underscores the need for patent applicants to carefully consider the context in which they are applying machine learning. The decision also emphasizes the importance of providing clear and concise explanations of the inventive concept.

Patent eligibility is determined by the Supreme Court in the case of Association for Molecular Pathology v. Myriad Genetics, Inc. In this case, the court held that patent eligibility is determined by the following criteria:

1. Is the subject matter directed to a law of nature, a natural phenomenon, or an abstract idea?
2. Is the subject matter directed to a process for a law of nature, a natural phenomenon, or an abstract idea?
3. Is the subject matter a means of a law of nature, a natural phenomenon, or an abstract idea?
4. Is the subject matter a product of a law of nature, a natural phenomenon, or an abstract idea?

Machine learning is a complex and rapidly evolving field, and its patentability is still a topic of debate.

Highlights

* The Federal Circuit emphasized the importance of demonstrating a technological improvement in machine learning claims. * The court rejected the argument that the application of machine learning to a new field of use confers patent eligibility. * The decision highlights the importance of considering the context in which machine learning is being applied. * The court observed that the patents in question relied on the use of generic machine learning technology in carrying out the claimed methods, and that the claimed machine learning technology was conventional.