CAFC Precedential Case of First Impression
On April 18, 2025, the U.S. Court of Appeals for the Federal Circuit (CAFC) decided a case of first impression regarding the intersection of patent claims directed to machine learning training and patentable subject matter under 35 U.S.C. § 101. The court stated, “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 under § 101.” (Recentive Analytics, Inc. v. Fox Corp., No. 23-2437 (Fed. Cir. Apr. 18, 2025))
In view of the court’s explicit holding, claims that recite solely training a generic model will not be found to be patent eligible under 35 U.S.C. § 101. Instead, machine learning training claims should recite limitations that describe a specific technical solution associated with a specific improved training process or a specific improved model.
What This May Mean for You
As new rejections under 35 U.S.C. § 101 are received, we recommend that pending machine learning training claims be reviewed so that the claims recite specific technical limitations regarding the training process. Likewise, we recommend that new patent applications that include claims directed to machine learning training be drafted with thorough support for such specific technical limitations.
While the Recentive Analytics decision should be of some concern to software-oriented businesses, the decision should have less of an impact on pending patent applications that are properly drafted. As new patent applications are drafted and prosecuted, any machine learning training claims should be reviewed by counsel to ensure that such claims are thoroughly supported by specific technical improvements to the training process or to the trained model.
[View source.]
