On January 10, 2018, the Federal Circuit added Finjan, Inc. v. Blue Coat Sys., Inc., No. 2016-2520 (Fed. Cir.), to its Enfish jurisprudence and upheld the subject matter eligibility of a software patent directed to virus-scanning downloadable app code for known and suspected malware. As construed, the invention claims novel behavioral-based analysis of source code to identify; detection of potentially dangerous files results in creation of a new file attached to the app code which is then evaluated by the destination computer to determine whether to allow the app to be downloaded.

Since the Supreme Court’s decision in Alice v. CLS Bank, 134 U.S. 2347 (2014), which applied the abstract idea judicial exception for §101 subject-matter eligibility to software patents, hundreds of software patents have been invalidated by district courts and the Federal Circuit. Alice’s now-familiar two-step approach determines first if a patent is directed to an abstract idea and second, if so, whether the claims are something more than a patent on the idea and thus patent-eligible. An abstract idea patent is eligible in Step 2 if it adds something more, namely a sufficiently inventive concept, to the abstract idea. In practice, it was almost a given that software patents satisfied Step 1 because they are necessarily directed to abstract ideas. Then, Enfish v. Microsoft, 822 F.3d 1327 (Fed.Cir. 2016), seemed to provide hope to a few software patentees in Step 1: claims relating to a database architecture that used a new, self-referential logical table were non-abstract at Step 1 because they focused on and provided “an improvement to computer functionality.” Since then, of the roughly 30 patentees claiming an improvement to computer functionality under Enfish, only four succeeded: McRO, Thales, Amdocs, and Visual Memory. Blue Coat is now the fifth and the Court did not even reach Step 2.

The method claim patent-eligible in Blue Coat consists of only three steps: (1) a virus-scanner (“inspector”) receives an app (“downloadable”); (2) the app code is scanned for recognizable malware and suspicious viruses (code that performs hostile operations) and a new file (“security profile”) is generated identifying dangerous code; and (3) the new file is linked to the app. Although it does not appear in the claimed steps, a computer receiving the app compares the security profile file to a “security policy” file on a user’s computer to determine whether the app should be accepted by a user’s computer. Virus-scanning is hardly new and had already been determined to be an abstract idea by the Federal Circuit – so, how did Finjan’s virus-scanning software survive Alice’s Step 1?

One distinguishing feature is its procedural posture: Blue Coat did not present an Alice challenge at the pleadings stage or in a summary judgment motion as have most Alice cases. Rather, it affirmed a district court’s post-trial denial of a JMOL motion. The standard of review was still de novo, but the Federal Circuit was faced with a fully litigated patent case, including claim constructions that significantly added to the few, very broad, very general terms used in the patent claims, where a jury had made a damages finding. And Finjan, the patentee, is a practicing entity that pioneered the patented technology. Although unspoken, these distinctions are relevant in the real world: the Court had the full benefit of claim construction and a contrary result would have overturned a completed trial and judgment won by a practicing entity. Nevertheless, such factors are likely not enough, singly or in combination, to explain the outcome.

A substantive explanation is that the Blue Coat panel attempted to expand the reach of Enfish both by relying on novelty in Step 1 and by conflating improved utility with improved functionality. In effect, the Federal Circuit created its own new two-step test for Alice Step 1’s abstract idea inquiry: first, whether the claimed method an advance over prior art; and second, whether the claimed advance improves computer functionality. With regard to the first step, the Blue Coat panel emphasized that the claimed “security profile” is an advance over prior art because it includes information about potentially hostile operations produced by a “behavior-based” virus scan of the app. The court distinguished this from traditional “code-matching” virus scans which recognize the presence of only previously-identified viruses contained in look-up tables. The Court’s analysis deviates from Alice’s claim-centric approach – “behavior-based” scanning is routed not in the claims themselves but in the specification disclosures relied on for claim construction. This method of virus scanning for code characteristics consistent with malware and then labelling potential virus code as suspicious is disclosed not in the claims but in the specification, and was a clear advance over prior art. For the second step, the Court asked whether the prior art advance constituted an improvement to computer functionality, and answered in the affirmative. The Blue Coat panel explained that the behavior-based, security profile approach protects against previously unknown viruses, and also “enables more flexible and nuanced virus filtering.”

While the method in Blue Coat, as construed, is arguably an improvement in the art and utility of malware detection, absent from the opinion is an explanation how this actually improved the functioning of a computer. The Court attempted to analogize the claim to that in Enfish, but the invention in Enfish was a new and unique improvement to the structure of the computer’s database memory that functioned differently than the prior art database. Specifically, the self-referential table of Enfish did not require “user[s to] adapt data entry” and provided “increased flexibility, faster search times, and smaller memory requirements.” Enfish, 822 F.3d 1327, 1337 (Fed. Cir. 2016). Behavior-based virus scanning does not provide such structural changes to computer memory or database architecture (virus-scanning is software) or improve the way a computer operates. Behavior-based scanning merely recognizes suspicious code based on the performance of actions (which themselves must be programmed into look-up tables) consistent with known virus programs.

Blue Coat represents a misapplication (or new extension) of Enfish and a deviation from post-Enfish decisions, including Bascom. Blue Coat attempts to distinguish only one of these cases, explaining that because the patented method attaches the virus scan results to the app in the form of a newly generated file, this is “a good deal more” than the virus-screening done on an intermediary computer held to be abstract in IV v. Symantec Corp, 838 F.3d 1307 (Fed. Cir. 2016). Again, though, the Court fails to explain how attaching a new file identifying suspicious code for the user’s computer to evaluate improves how the scanning computer functions versus improving the scope of virus detection. Scanning code and creating files are typical, expected functions of computers; reading code and recognizing real and suspected malware is a typical, expected function of software. Moreover, all software adds functionality and runs on a computer.

The net effect of Enfish, Blue Coat, and related cases is that the Federal Circuit effectively combines the two-part Alice test into its own permutation of Step 1 for a specific class of software patents – those claiming to be directed to improving the functionality of a computer.  The “advancement over the prior art” part of the Blue Coat test turns on whether the claim includes specific language defining an invention that is not conventional. It is unclear how this test would be applied at the pleadings stage where subject-matter eligibility is limited to the pleadings. Also, inquiry is not substantially different than Alice’s Step 2 analysis, but under Blue Coat if the claimed invention survives Alice Step 1, there is no need to proceed to Step 2.

The inquiries undertaken by Blue Coat’s new two-step Step 1 methodology are rooted in the Federal Circuit’s own cases rather than Supreme Court jurisprudence. Focus on the “claimed advance” over the prior art at Step 1 has its foundation in Genetic Techs. v. Merial, 818 F.3d 1369 (Fed. Cir. 2016), while an “improvement to computer functionality” is rooted in Enfish. Critics will note the fine line of looking at prior art in Step 1, because advancements over prior art are traditionally reserved for inquires under 35 U.S.C §§ 102 and 103 – not § 101. What if a court concludes that the first element of the Blue Coat test is met, but the second (improvement to computer functionality) is not?  Arguably, the court would then have to proceed to Alice Step 2, but it is difficult to perceive how Step 2 would not necessarily be satisfied if the claimed invention was already viewed as an advancement over prior art.

Blue Coat presents several challenges for practitioners advising practicing entities on subject-matter eligibility of asserted software patents. Post-Blue Coat, how does a district court distinguish an improvement to computer functionality from software that adds utility by merely using a computer in the way it is meant to be used? Ironically, lower courts criticized the Supreme Court’s so-called failure to provide clearer guidance in Alice. Some critics seek further guidance from the Supreme Court, others seek a “fix” from Congress. Unfortunately, the Federal Circuit’s attempts to add clarity in its Enfish software jurisprudence suffer from the same defect regarding the meaning of “computer functionality.” Blue Coat may also embolden non-practicing entities to renew assertions of patents which present close calls – especially where specificity lacking in claims can be found in the specification. Ultimately, the decision may warrant review by the Federal Circuit en banc or even the Supreme Court, because panel attempts at clarification are missing the mark. For more information on this issue, please contact Erin Knese or Steve Holtshouser.