On April 18, 2025, the United States Court of Appeals for the Federal Circuit affirmed the dismissal of a patent infringement suit brought by Recentive Analytics, Inc. against Fox Corporation. See Recentive Analytics, Inc. v. Fox Corp., No. 23-2437 (Fed. Cir. Apr. 18, 2025). Recentive had sued Fox on four patents comprising two categories: the “Machine Learning Training” patents (U.S. Patent Nos. 11,386,367 and 11,537,960), which described methods for scheduling live events using machine learning; and the “Network Map” patents (U.S. Patent Nos. 10,911,811 and 10,958,957), which addressed the optimization of broadcaster programming schedules across geographic markets. Slip Op. at 3, 5. Fox moved to dismiss for failure to state a claim on the ground that the patents are ineligible under § 101. Id. at 7. The district court granted the motion finding that Recentive’s patents were directed to ineligible subject matter under 35 U.S.C. § 101. Id. at 9.
On appeal, the Federal Circuit applied the familiar two-step test for subject matter eligibility under Alice Corp. v. CLS Bank Int’l, 573 U.S. 208 (2014). Id. at 9-10. At step one, the Court asked whether the claims were directed to an abstract idea. Id. At step two, the Court considered whether the claims contained an “inventive concept” sufficient to transform the abstract idea into patent-eligible subject matter. Id.
At step one of the Alice inquiry, the Federal Circuit concluded that the focus of Recentive’s claimed invention, as a whole, was directed to an abstract idea. Id. at 10. The Federal Circuit first observed that Recentive had repeatedly conceded that it was not claiming any novel machine learning techniques. Id. at 11. Both the “Machine Learning Training” patents and the “Network Map” patents simply employed generic machine learning technology in the context of generating event schedules and network maps. Id.
The Federal Circuit rejected the argument that features like iterative training or real-time dynamic updating rendered the claims non-abstract. It noted that such features are inherent to the very nature of machine learning itself, and thus did not reflect an improvement to the underlying technology. Id. at 12. Recentive’s own briefing and hearing statements acknowledged that iterative training and updating are conventional aspects of machine learning processes. Id.
Moreover, the Federal Circuit scrutinized whether the claims disclosed how the purported results were achieved. It found that the claims merely stated functional objectives—optimizing schedules or network maps—without describing how an improvement was accomplished. Id. at 13.
The Federal Circuit distinguished Recentive’s patents from claims found eligible in Enfish, LLC v. Microsoft Corp., 822 F.3d 1327 (Fed. Cir. 2016), and Koninklijke KPN N.V. v. Gemalto M2M GmbH, 942 F.3d 1143 (Fed. Cir. 2019). Unlike those cases, which involved a specific implementation of a solution to a problem in the software arts or a specific means or method that solves a problem in an existing technological process, Recentive’s patents merely applied generic machine learning to a new environment—namely, scheduling and network mapping. Id. at 13.
The panel further noted that merely applying an abstract idea to a new technological environment or field of use does not transform it into patent-eligible subject matter. Id. at 14. The Court found that scheduling events and generating network maps were long-practiced human activities and that the introduction of machine learning did not alter the basic abstract character of the claims. Id.
Additionally, the Court rejected the proposition that improving the speed and efficiency of these manual tasks through the use of conventional machine learning techniques sufficed to render the claims patent-eligible. As the Court explained, generic improvements in speed and efficiency arising from automation are insufficient to confer eligibility under § 101. Id. at 15–16.
At step two, the Court found that the patents lacked an inventive concept that would render the abstract ideas patent eligible. The only asserted inventive concept was using machine learning to dynamically generate optimized maps and schedules based on real-time data and update them based on changing conditions. Id. at 16. But this inventive concept amounted to nothing more than claiming the abstract idea itself. Id. The Court found nothing in the claims, whether considered individually or in their ordered combination, that would transform the Machine Learning Training and Network Map patents into something “significantly more” than the abstract idea of generating event schedules and network maps through the application of machine learning. Id. at 17.
In short, the Federal Circuit affirmed the district court’s dismissal under Rule 12(b)(6), holding that all four asserted patents were invalid under § 101. The Court found the claims to be directed to abstract ideas without the requisite inventive concept and further held that amending the complaint would have been futile.
The Recentive decision represents the Federal Circuit’s first-ever patent eligibility decision involving machine learning. The Federal Circuit confirmed that merely using artificial intelligence (AI) or machine learning to make a task faster or more efficient does not make an invention patent eligible under 35 U.S.C. § 101. The Court emphasized that applying conventional machine learning methods to new data environments without technological innovation constitutes an abstract idea, not a patent-eligible invention.
This ruling has particular significance for the burgeoning field of AI. Patent applicants seeking protection for machine learning-based innovations must ensure their claims do more than merely apply conventional algorithms to new datasets. The Federal Circuit’s opinion reiterated that patent eligibility hinges not on the newness of the field of use, but on the technological substance of the claimed advance. Although the Federal Circuit suggested that true technological improvements in AI might qualify for patent protection, it did not provide concrete guidance on what such improvements would look like.
As courts begin to grapple with the evolving interface between AI and intellectual property, Recentive serves as a cautionary touchstone that the same fundamental eligibility principles will apply for those seeking to protect AI-based inventions.
