On Friday, April 18, 2025, the Federal Circuit addressed a question of first impression regarding the validity of certain machine-learning patents under Section 101 in Recentive Analytics, Inc. v. Fox Corp., et al., 2023-2437. Specifically, the court opined on “whether claims that do no more than apply established methods of machine learning to a new data environment are patent eligible” under Section 101.
The challenged patents fell into two groups. The first used machine learning to optimize live-event schedules based on a user’s target features (e.g., maximizing event attendance, revenue, or ticket sales) and iteratively updated that schedule with real-time changes in data. The second used machine learning to optimize network maps, which determine the programs or content displayed by a broadcaster’s channels within certain geographic markets at particular times.
Applying the two-step Alice framework familiar in software patent cases, the Federal Circuit affirmed the lower court’s holding of subject matter ineligibility “because the patents are directed to the abstract idea of using a generic machine learning technique in a particular environment, with no inventive concept.” But, recognizing that “[m]achine learning is a burgeoning and increasingly important field and may lead to patent-eligible improvements in technology,” the court emphasized the limited nature of its holding: “[t]oday, 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.”
For Alice Step 1, the court held that it was “clear” that the disputed claims were directed to ineligible, abstract subject matter. As an initial matter, Recentive admitted that its patents did not claim machine learning itself but rather its application to event schedules and network maps. The specifications for both sets of patents were notable in this regard with both teaching that the patent claims employ “any suitable machine learning technique.” The court observed that “[b]oth sets of patents rely on the use of generic machine learning technology in carrying out the claimed methods for generating event schedules and network maps,” and the claimed machine learning technology was “conventional.”
The Federal Circuit 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 “[a]n abstract idea does not become nonabstract by limiting the invention to a particular field of use or technological environment.”
Searching then for a technological improvement, the court uncovered none. The claimed iterative training and dynamic adjusting in the machine learning model could not constitute a technological improvement as such functions are “incident to the very nature of machine learning.” Further, the greater speed and efficiency gained through the application of machine-learning to the tasks of event scheduling and network mapping could not confer patent eligibility.
Recentive argued that its claimed application of machine-learning was not generic because Recentive had found a way to make its algorithms function dynamically and uncover previously unrecognized patterns in the data. But the court then pointed to Recentive’s other concession: the patents did not claim a specific method for “improving the mathematical algorithm or making machine learning better.” Still, the court’s decision did not rest exclusively on Recentive’s admissions. It observed that “the claims do not delineate steps through which the machine learning technology achieves an improvement.”
For Alice Step 2, Recentive argued that the inventive concept was the use of machine learning to dynamically generate and update optimized maps and schedules based on real-time data. But the Federal Circuit observed that “this is no more than claiming the abstract idea itself.” The court found nothing in the claims, individually or in their ordered combination, that would transform the patents into something significantly more than “the abstract idea of generating event schedules and network maps through the application of machine learning.”
Lastly, the court rejected Recentive’s argument that it should have been granted leave to amend because “Recentive failed to propose any amendments or identify any factual issues that would alter the § 101 analysis.”
Takeaways:
- 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, claim patent ineligible subject matter under Section 101.
- The Federal Circuit’s Alice analysis for machine language models differs little from its applications in the more traditional software contexts. The Federal Circuit also notably states that its cases holding that “the application of existing technology to a novel database does not create patent eligibility” are apropos to the machine learning context, citing its decisions in SAP Am., Inc. v. InvestPic, LLC, 898 F.3d 1161 (Fed. Cir. 2018) and Power Grp., LLC v. Alstom S.A., 830 F.3d 1350, 1356 (Fed. Cir. 2016), among others. Practitioners may want to consider these decisions in composing their Section 101 arguments.
