Recentive Analytics, Inc. v. Fox Corp
Authored by: Jeremy J. Gustrowsky
In a recent decision, the Federal Circuit tackled whether simply using machine learning in a new context—like TV scheduling or network mapping—can make an invention eligible for a patent. Recentive Analytics, Inc. owned four patents (U.S. Patent Nos. 10,911,811; 10,958,957; 11,386,367; and 11,537,960) that described using generic machine learning models to generate event schedules and network maps for television broadcasts and live events. Recentive sued Fox Corp. and its affiliates for allegedly infringing these patents, but the district court dismissed the case, finding the patents were not eligible for protection under U.S. patent law.
On appeal, the Federal Circuit agreed with the lower court. The judges explained that just applying existing machine learning techniques to a new type of data or industry—without improving the machine learning technology itself—does not make an invention patent-eligible. The court pointed out that the patents did not describe any new or inventive way of using machine learning; they simply used off-the-shelf methods like neural networks or decision trees to automate tasks that humans had previously done manually, such as creating TV schedules or network maps.
The court emphasized that using computers or machine learning to do something faster or more efficiently than a human does not, by itself, make the process patentable. For a software or machine learning invention to be eligible, it must offer a specific technological improvement or a new way of doing something—not just automate a known process in a new setting. The court also noted that the patents in question only described the idea of using machine learning in these fields, without explaining any new way to make the technology work better.
This decision serves as a reminder: simply applying a generic technology like machine learning into a new area isn’t enough for a patent. Inventors need to show a real technological advance or improvement to the underlying technology to clear the eligibility hurdle.