Recentive Analytics v. Fox Corp., decided April 18, 2025

Date: April 18, 2025

Subject: Analysis of the Federal Circuit's Decision on Patent Eligibility of Machine Learning for Broadcast Scheduling

Executive Summary:

The Federal Circuit Court of Appeals affirmed the District Court's dismissal of Recentive Analytics, Inc.'s patent infringement lawsuit against Fox Corp. The court held that Recentive's four patents, directed to using machine learning for generating network maps and event schedules for television broadcasts and live events, are directed to patent-ineligible subject matter under 35 U.S.C. § 101. The court reasoned that the patents claim the abstract idea of applying generic machine learning techniques to a particular field without any inventive concept or improvement to the machine learning technology itself. This decision establishes that merely applying existing machine learning methods to new data environments, without more, does not confer patent eligibility.

Main Themes and Important Ideas/Facts:

The Patents in Question:

Recentive owned four patents: U.S. Patent Nos. 10,911,811 ('811 patent), 10,958,957 ('957 patent), 11,386,367 ('367 patent), and 11,537,960 ('960 patent).
These patents fall into two categories:
Machine Learning Training Patents ('367 and '960): Focused on dynamically generating event schedules using machine learning to optimize based on event parameters and target features.
Claim 1 of the '367 patent is representative, outlining steps for collecting data, iteratively training an ML model (neural network or support vector), outputting an optimized schedule, and updating it based on real-time changes.
The specification notes that "the machine learning model may be ‘trained using a set of training data,’ which can include ‘historical data from previous live events or series of live events.’"
It explicitly states that "the patented method employs ‘any suitable machine learning technique[,] . . . such as, for example: a gradient boosted random forest, a regression, a neural network, a decision tree, a support vector machine, a Bayesian network, [or] other type of technique.’"
Network Map Patents ('811 and '957): Focused on automatically and dynamically generating network maps for broadcasters, optimizing television ratings.
Claim 1 of the '811 patent is representative, describing receiving broadcasting schedules, generating a network map using a machine learning technique to optimize overall television ratings, automatically updating the map, and using it to determine program broadcasts.
Training data for the network map patents could include "weather data, news data, and/or gambling data."
Similar to the other patents, the specification allows for the use of "any suitable machine learning technique."

The District Court's Ruling:

The District Court granted Fox's motion to dismiss, finding the patents ineligible under the two-step Alice Corp. v. CLS Bank International framework.
The court concluded that the claims were "directed to the abstract ideas of producing network maps and event schedules, respectively, using known generic mathematical techniques."
At the second step of Alice, the District Court found no "inventive concept" because the machine learning limitations were "broad, functionally described, well-known techniques" using "only generic and conventional computing devices."

The Federal Circuit's Affirmation:

The Federal Circuit reviewed the dismissal de novo and affirmed the District Court's decision.
Step One of Alice (Directed to an Abstract Idea):
The court emphasized that the focus is on "the specific asserted improvement in computer capabilities... or, instead, on a process that qualifies as an abstract idea for which computers are invoked merely as a tool."
The court noted Recentive's concession that it was not claiming the machine learning techniques themselves, but rather their application to specific contexts.
The Federal Circuit found that both sets of patents rely on "generic machine learning technology" and that the described technology is "conventional," citing the patents' own specifications listing various well-known machine learning techniques.
The court dismissed Recentive's argument that "iterative training" or "dynamic adjustments" represented a technological improvement, stating that these are "incident to the very nature of machine learning." The court quoted Recentive's own arguments: "'using a machine learning technique[] . . . necessarily includes [an] iterative[] training step . . . .'" and "'[T]he way machine learning works is the inputs are defined, the model is trained, and then the algorithm is actually updated and improved over time based on the input.'"
The court rejected Recentive's argument about dynamically functioning algorithms and unearthing previously unrecognizable patterns, noting Recentive's admission that the patents do not claim a specific method for "improving the mathematical algorithm or making machine learning better."
The court emphasized that the claims did not describe how any improvement was achieved, stating, "That is, the claims do not delineate steps through which the machine learning technology achieves an improvement."
The Federal Circuit reiterated the established principle that "[a]n abstract idea does not become nonabstract by limiting the invention to a particular field of use or technological environment."
The court also held that "the application of existing technology to a novel database does not create patent eligibility."
Finally, the court stated that the increased speed and efficiency resulting from using computers with existing machine learning techniques do not, by themselves, create patent eligibility. The court cited cases like Content Extraction and DealerTrack.
Step Two of Alice (Inventive Concept):
The court found that Recentive's assertion that the 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" was merely claiming the abstract idea itself.
The Federal Circuit concluded that there was nothing in the claims, individually or in combination, that transformed the abstract idea into a patent-eligible application.

Denial of Leave to Amend:

The Federal Circuit upheld the District Court's denial of Recentive's request for leave to amend, finding that further amendment would be futile as Recentive failed to propose any amendments or identify factual issues that would alter the § 101 analysis.

Claim Construction at the Pleading Stage:

The court briefly addressed Recentive's suggestion that the District Court erred in resolving claim-construction disputes at the pleading stage, finding no error because Recentive had failed to identify claim terms requiring construction that could affect the patent-ineligibility analysis. The court cited Trinity Info Media, LLC v. Covalent, Inc. in support.

Key Quotes:

"We affirm because the patents are directed to the abstract idea of using a generic machine learning technique in a particular environment, with no inventive concept."
"This case presents a question of first impression: whether claims that do no more than apply established methods of machine learning to a new data environment are patent eligible. We hold that they are not."
"Recentive has repeatedly conceded that it is not claiming machine learning itself."
"Iterative training using selected training material and dynamic adjustments based on real-time changes are incident to the very nature of machine learning."
"'using a machine learning technique[] . . . necessarily includes [an] iterative[] training step . . . .'" (Quoting Recentive's Opposition Brief)
"[A]n abstract idea does not become nonabstract by limiting the invention to a particular field of use or technological environment."
"We have also held the application of existing technology to a novel database does not create patent eligibility."
"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."

Conclusion:

The Federal Circuit's decision in Recentive Analytics v. Fox Corp. reinforces the stringent standards for patent eligibility, particularly in the realm of software and machine learning. The ruling clarifies that simply applying existing machine learning techniques, even to novel data or problem spaces, does not automatically render an invention patent-eligible. To achieve patent eligibility, claims involving machine learning likely need to demonstrate a specific improvement in the machine learning technology itself or a non-conventional application that goes beyond the inherent functionality of generic machine learning processes. This case serves as an important precedent for assessing the patent eligibility of future inventions leveraging machine learning in various industries. Full opinion