The Federal Circuit in a 2-1 decision upheld four software patents against a patent-eligibility challenge, finding that the patents do not claim an “abstract idea.” The decision, Amdocs (Israel) Ltd. v. Openet Telecom Inc. et al.,  makes clear that courts will evaluate patent eligibility, under 35 U.S.C. § 101 and the Supreme Court’s Alice decision, using an individual, case-by-case approach: courts will compare the challenged patents and claims to those from previous cases in a traditional common-law style.
Alice’s software patents covered settling financial transactions via a third-party intermediary, and the Court unanimously held them invalid under § 101. 
The Alice decision, and its now-familiar two-step test, created substantial uncertainty about the future of software patents especially since the Court stated that “[t]he introduction of a computer into the claims does not alter the analysis at . . . step two.” Would any survive? If some did, which ones would and why? Since Alice, the Federal Circuit has decided several cases involving § 101 challenges to software patents—upholding some software patents, invalidating others. Enter, Amdocs.
Broadly speaking, Amdocs’s patents cover “parts of a system designed to solve an accounting and billing problem faced by network service providers,” specifically the “massive record flows” and requirement for “huge databases” in large networks. Amdocs’s patented system’s components “are arrayed in a distributed architecture,” which “enable load distribution” and “reduc[e] congestion in network bottlenecks while still allowing data to be accessible from a central location.”
Amdocs sued Openet in 2010 for infringement of four of its patents. During Amdocs’s first appeal on claim construction, the Supreme Court decided Alice. On remand, the district court held that Amdocs’s patents were directed to patent-ineligible subject matter under § 101 in light of Alice. Amdocs then filed its second appeal.
The Amdocs Court’s Opinion
The Amdocs majority reversed the district court, holding that Amdocs’s patents were patent-eligible under Alice step two. The majority suggested that Amdocs’s patents also passed Alice step one, but the opinion is unclear. (The dissent did not think that the majority engaged in Alice step one at all.)
At the outset, the majority noted that Alice’s two steps often overlap. Although step two usually involves “more comprehensive analysis,” the court noted that “in some situations this analysis could be accomplished without going beyond step one.”
The majority explicitly refused to articulate a “single, universal definition of an ‘abstract idea,’” noting the difficulty in fashioning a definition for “as-yet-unknown cases with as-yet-unknown inventions.” The majority observed that the “decisional mechanism courts now apply” is simply to look at earlier cases with a “similar or parallel descriptive nature”—i.e., similar fact patterns.
The court embraced this “classic common law methodology,” especially for determining whether a patent covers an “abstract idea.” The majority defended its “more flexible” approach as being the one that the Supreme Court uses. For added support, the majority also cited the U.S. Patent and Trademark Office’s post-Alice Examination Policy—a policy that instructs examiners that, “when performing an analysis of whether a claim is directed to an abstract idea . . ., examiners are to continue to determine if the claim recites (i.e., sets forth or describes) a concept that is similar to concepts previously found abstract by the courts.” 
The majority gave some guidance about how courts should perform this common-law approach. Courts should discuss “only the most relevant prior opinions, rather than every prior opinion in an actively litigated field,” as this is a “necessary discipline.”
The Amdocs majority applied this common-law approach and held that all four patents’ claims were patent eligible and not directed to “abstract ideas.” The court compared the claims in Amdocs’s patents to five prior cases—two in which the claims were upheld,  and three in which the claims were struck down. 
Like the former two cases, Amdocs’s claims “solve a technological problem” and “improve the performance of the system itself.” And unlike the latter three cases, Amdocs’s claims are “narrowly drawn to not preempt any and all generic enhancement of data in a similar system,” and do not “merely combine the components in a generic manner.”
In particular, the court relied on a data-enhancing limitation in the claims of Amdocs’s patents. In Amdocs’s “distributed architecture,” or decentralized network configuration, user data collected from a single source lacks the requisite billing and accounting information, so the data is “enhanced”: Amdocs’s system combines data from other sources and user sessions to create “meaningful session records” for the service provider.
Judge Reyna dissented. In his view, “the abstract idea exception must be applied in a way that reserves patent protection for means rather than for ends,” so that the patent system maintains incentives to develop “new ways of achieving the same result more cheaply and efficiently.” “[A] desired goal without means for achieving that goal is an abstract idea.”
Judge Reyna gave particular weight to the “directed to” language from Alice for step one. For him, “directed to” must mean “more than merely ‘embody, use, reflect, rest upon, or apply.’” If the claim does not describe how a particular goal is accomplished, it likely is directed to an abstract idea. But if the claim “recites specific structure or function for accomplishing the goal,” then the claim is more likely patent eligible.
Judge Reyna would further employ this reasoning at Alice step two: “[c]laims that fail to recite how a desired goal is accomplished do not recite an inventive concept.” Using this mode of analysis, he would have held the claims of two of the four patents patent-eligible, and the other two ineligible. He did not find Amdocs’s “enhancement” argument persuasive both because some claims did not provide context for the enhancement limitation and because the claims recited “no components or structure” for the enhancement. But other claims included sufficient structural limitations—such as “a plurality of network devices at a plurality of layers”—that, in Judge Reyna’s view, claimed how Amdocs’s distributed architecture worked, thereby making those claims patent eligible.
In his dissent, Judge Reyna criticized the majority for “declining to engage in the step 1 inquiry,” citing numerous post-Alice Federal Circuit cases. He also criticized the majority for relying on the “distributed architecture” of Amdocs’s patents because that phrase does not appear in the claim language. “If I were to examine only the written description of the asserted patents, I would conclude that the network monitoring system disclosed therein is eligible for patenting. . . . But the inquiry is . . . whether the claims are directed to a patent eligible concept.”
The § 101 landscape will continue to evolve post-Alice. Clarity will come with time: “In the complexities of § 101, the law is evolving into greater certitude based on experience, not on generalizations.”