Two 2026 Federal Circuit rulings show why Section 101 eligibility depends on what the claim requires, not what the specification discloses.

Under § 101, patent eligibility generally cannot depend on a technical improvement that the asserted claim does not require. Two Federal Circuit decisions in 2026 illustrate this principle in different ways: Constellation Designs, LLC v. LG Electronics Inc. addresses claims that recite a desired technical result without requiring how to achieve it, while Berkeley*IEOR v. W.W. Grainger Inc. addresses an asserted improvement that the claim did not require at all.

In Constellation Designs, the Federal Circuit addressed the issue in a precedential decision involving digital communications technology. The opinion was issued on April 28, 2026 and modified and reissued on August 31, 2026 following a petition for rehearing. The modification did not change the § 101 analysis discussed below.

More recently, in BerkeleyIEOR, decided September 21, 2026, the Federal Circuit addressed a related issue involving computer-implemented profitability calculations. BerkeleyIEOR is expressly designated nonprecedential and therefore does not carry the precedential weight of Constellation Designs. Its reasoning, however, expressly relies on the Federal Circuit’s existing § 101 precedent concerning unclaimed technical features.

Constellation Designs
: Result-Oriented Optimization Versus a Specific Constellation
The patents in Constellation Designs concerned digital communication systems using signal constellations. The Federal Circuit divided the asserted claims into two groups: “optimization claims” and “constellation claims.”

Claim 17 of U.S. Patent No. 8,842,761 was representative of the optimization claims. Its final limitation required “the QAM symbol constellation [to be] a geometrically spaced symbol constellation optimized for capacity using parallel decode capacity that provides a given capacity at a reduced signal-to-noise ratio compared to a QAM signal constellation that maximizes dmin.”

The Federal Circuit focused directly on what this language did and did not require:

“The claim itself does not recite how to achieve a constellation ‘optimized’ for PD capacity. Rather, this important and distinguishing element of the claim is described in a result-oriented way.”

The court concluded that “claim 17 is ineligible because it is an abstract, result-oriented claim directed to all ways of achieving a recited result.”

The specification contained additional technical disclosure. It described, for example, an iterative process involving selection of a constellation size and desired capacity, an initial SNR estimate, optimization of the constellation at that SNR, determination of an output SNR, and further iterations until convergence.

But the Federal Circuit identified the relationship between the specification and claim language expressly “although ‘the specification may help illuminate the true focus of a claim, when analyzing patent eligibility, reliance on the specification must always yield to the claim language.'”

The court further explained that claim 17 lacked sufficient details concerning the asserted “concrete technique.”

Claim 21  Provides the Counterexample
The same decision provides a useful comparison because a different group of claims was held eligible.

Claim 21 of U.S. Patent No. 11,019,509 required a symbol constellation in which “the plurality of unique point locations are unequally spaced;” “the constellation points each have a location and a different label; and “the locations of at least two of the constellation points are the same.”

The Federal Circuit expressly contrasted claim 21 with the optimization claims by stating that “[w]e conclude that this distinction between the result-oriented optimization claims and constellation claims makes all the difference.”

The court explained that, rather than broadly claiming all means of optimizing, claim 21 applied the disclosed process using specified parameters. It then identified both the technological problem and the claimed solution: “claim 21 in its entirety is directed to solving a particular technological problem—overcoming capacity constraints to improve coding gains—using a particular technological solution—specific, non-uniform constellations with overlapping constellation point locations.”

The court therefore held “the limitations in claim 21 provide enough specificity and structure to satisfy Alice step one.”

Claim 21 was patent eligible, and the Federal Circuit did not proceed to Alice step two for that claim.

Thus, Constellation Designs provides examples on both sides of the eligibility analysis within closely related technology: representative claim 17 recited an optimization in terms the court considered result-oriented, while representative claim 21 expressly recited particular characteristics of the constellation that the court found sufficiently specific and structured.

Berkeley*
IEOR: The Asserted Improvement Was Not Required by the Claim
The Federal Circuit addressed a related issue at Alice step two in Berkeley*IEOR v. W.W. Grainger Inc.

The patents concerned methods for calculating object-level profitability. Claim 1 of U.S. Patent No. 7,596,521 was representative. Among other limitations, claim 1 required “using the relational database management system to independently calculate at least one marginal value of profit for each object being measured using the established rules as applied to a selected set of prepared information.”

The district court found an inventive concept based on an asserted improvement in computer operation through faster parallel computations.

The claim, however, did not require parallel processing.

The district court’s claim construction did not require parallel processing, leaving room for the calculations to be performed sequentially—or not at all. On appeal, Berkeley’s counsel also conceded that parallel processing “[is] not a claim requirement.”

The Federal Circuit identified the controlling issue directly “if a claim is so broad that it can be practiced without the alleged inventive concept, i.e., the alleged inventive concept is not required by the claim, then the claim cannot be ‘significantly more’ than the abstract idea itself and fails at step two.”

The court then restated the point in terms of an asserted technological improvement: “when the invention as claimed does not necessitate the alleged ‘improvement in . . . [the] technology or technical field,’ then such alleged improvement cannot save the claim at step two.”

Applying that principle, the Federal Circuit held that it was legal error for the district court to rely on parallel processing as the inventive concept because a system that did not perform calculations in parallel could still fall within the scope of claim 1.

The Federal Circuit reversed-in-part and affirmed-in-part: it reversed the district court’s summary judgment of patent eligibility and affirmed judgment for the defendants on the alternative ground that the asserted claims are patent ineligible under § 101.

The Broader Line of Federal Circuit Reasoning
Berkeley*IEOR is nonprecedential, but its reasoning did not purport to create a new rule. The panel expressly connected its conclusion to existing Federal Circuit precedent.

Quoting its 2026 precedential decision in Trustees of Columbia University v. Gen Digital Inc., the court stated that claims cannot be “directed to a technological improvement when nothing in the claims requires the steps necessary to make the improvement.”

The Berkeley*IEOR panel also cited GoTV Streaming, LLC v. Netflix, Inc. for the proposition that “only features that are claimed, not unclaimed details that appear in the specification, can supply something beyond . . . an abstract idea and sufficient to render the claim eligible.”

Constellation Designs expresses a related principle at Alice step one. Citing ChargePoint, the precedential opinion states “although ‘the specification may help illuminate the true focus of a claim, when analyzing patent eligibility, reliance on the specification must always yield to the claim language.'”

And at Alice step two, Constellation Designs quotes Two-Way Media: “[t]o save a patent at step two, an inventive concept must be evident in the claims.”

Together, these statements support the article’s central point: an eligibility theory based on a technical improvement generally must be tied to limitations that the asserted claim actually requires.

Patent-Drafting Considerations
The decisions do not establish a rule that every technical detail disclosed in a specification must be incorporated into an independent claim. Nor do they prescribe how narrowly a claim must be drafted to satisfy § 101.

They do, however, provide a basis for a practical drafting consideration.

When patent eligibility may depend on a particular technical implementation or improvement, practitioners should consider whether the relevant claim actually requires the feature on which that eligibility position would depend.

Constellation Designs provides a particularly useful illustration. The Federal Circuit found representative claim 17 result-oriented where the claim recited an optimized constellation without reciting how the optimization was achieved. By contrast, representative claim 21 expressly required particular constellation characteristics, including unequally spaced point locations and overlapping point locations, and the court held that those limitations provided “enough specificity and structure to satisfy Alice step one.”

Berkeley*IEOR presents the issue differently. The asserted improvement was parallel processing, but the representative claim did not require parallel processing. The Federal Circuit therefore held that parallel processing could not supply the inventive concept at Alice step two.

For preparation and prosecution of software and other computer-implemented inventions, these decisions support reviewing the claims against the technical-improvement theory that may ultimately be needed under § 101. A specification can and should disclose appropriate technical detail and alternative implementations. But if eligibility is expected to depend on a particular technical feature, a separate question remains: does the claim being relied upon actually require that feature?

Cases

Constellation Designs, LLC v. LG Electronics Inc., No. 2024-1822 (Fed. Cir. Apr. 28, 2026; modified and reissued Aug. 31, 2026) (precedential).

Berkeley*IEOR v. W.W. Grainger Inc., Nos. 2024-2353, 2025-1009 (Fed. Cir. Sept. 21, 2026) (nonprecedential).

Trustees of Columbia University in the City of New York v. Gen Digital Inc., 169 F.4th 1320 (Fed. Cir. 2026).

GoTV Streaming, LLC v. Netflix, Inc., 166 F.4th 1053 (Fed. Cir. 2026).

Disclaimer: This article is for general informational purposes only and does not constitute legal advice or create a patent practitioner-client relationship.