In the nearly three years since Alice Corp. v. CLS Bank Int’l, 134 S. Ct. 2347 (2014), the Federal Circuit has analyzed the issue of patent-eligibility in the context of numerous distinct technologies. The court’s precedent applying Alice addresses patents directed to financial services, Internet content filtering, database maintenance, payment-based content access, animated facial gesturing, physical object tracking, distributed data processing, and more. One takeaway is that no technological discipline is, per se, eligible or ineligible. Instead, the outcomes of these cases are better explained by the specific inventions at issue and how they are claimed. RecogniCorp, LLC v. Nintendo Co., Ltd., No. 2016-1499 (Fed. Cir. Apr. 28, 2017), confirms this lesson.

The asserted patent in RecogniCorp was directed to a method and apparatus for building a composite facial image using constituent parts. One problem that the patent attempted to address was the tradeoff between file size and resolution of composite facial images: higher resolution images require more space in memory. The patent thus sought to encode an image at one point through a variety of image classes that required less memory and bandwidth, and decode the image at another point. Independent claim 1, which the Federal Circuit considered “representative” on appeal, applied these concepts to display facial feature images, select a facial feature image associated with a code, perform a multiplication operation on the code, and reproduce the composite image based on the code.

RecogniCorp, the assignee of the patent, sued Nintendo for alleged infringement in the District of Oregon. The patent thereafter survived a reexamination challenge, which resulted in amendments to all independent claims relating to the requirement of performing “multiplication operations” on facial codes. Nintendo responded by filing a motion for judgment on the pleadings, arguing that the claims were patent-ineligible under 35 U.S.C. § 101. The district court granted the motion without conducting claim construction. According to the district court, the claims were “directed to the abstract idea of encoding and decoding composite facial images using a mathematical formula” and added no inventive concept beyond that abstract idea. The district court further noted that the process of selecting images associated with codes, and applying corresponding operations, was similar to a “paint by numbers” game.

The Federal Circuit affirmed the district court’s invalidity finding on appeal. First, before engaging in the two-step Alice analysis, the Federal Circuit reaffirmed that “not all claims in all software patents are necessarily directed to an abstract idea.” Slip op. at 6. Nevertheless, the particular claims in this case were “directed to the abstract idea of encoding and decoding image data.” Id. at 7. As the Federal Circuit explained, the claimed technique merely used standard encoding and decoding of images, analogous to “Morse code, ordering food at a fast food restaurant via a numbering system, and Paul Revere’s ‘one if by hand, two if by sea’ signaling system.” Id. The Federal Circuit further contrasted the patent’s use of a “mathematical formula” with the invention in Diamond v. Diehr, 450 U.S. 175 (1981), because the claims in Diehr involved more than a mere abstract concept (mathematics), whereas the claims in RecogniCorp’s patent merely added that abstract concept to another—namely, encoding and decoding. The Federal Circuit also found its earlier (pre-Alice) decision in Digitech Image Techs., LLC v. Elecs. for Imaging, Inc., 758 F.3d 1344 (Fed. Cir. 2014), instructive. That case involved claims covering the abstract idea of organizing information through mathematical correlations. According to the Federal Circuit, both Digitech and RecogniCorp’s patent involved “[a] process that started with data, added an algorithm, and ended with a new form of data was directed to an abstract idea.” Slip op at 9.

Agreeing with the district court on the abstract idea inquiry, the Federal Circuit next considered RecogniCorp’s arguments under Alice’s step two. The Federal Circuit found that the claims recited nothing substantially more than the concept of encoding and decoding image data. Indeed, the Federal Circuit noted that the presence of a mathematical algorithm in the claims failed to supply an inventive concept. Instead, the algorithm simply changed data into other forms of data. Further undercutting RecogniCorp’s case for an inventive concept, the Federal Circuit noted that claim 1 of the asserted patent did not actually require the use of a computer, and while other claims did require a computer, they just mirrored the abstract idea recited in claim 1 using generic computer components.

In view of RecogniCorp and the numerous other post-Alice Federal Circuit decisions, patent practitioners now have ample precedent to consider in evaluating the patentability of particular inventions and claims. While the answer on patentability may not always be clear, RecogniCorp and the Federal Circuit’s other cases do confirm that there are no bright-line rules that render entire technological fields eligible or ineligible for patenting. For example, while RecogniCorp dealt with a patent for image analysis that was allegedly infringed by a video gaming system manufacturer, McRO, Inc. v. Bandai Namco Games Am. Inc., 837 F.3d 1299 (Fed. Cir. 2016), involved a patent for animation analysis that was allegedly infringed by video game developers (which was not found ineligible). Despite the surface similarity between the technical fields of such patents, the specifics of the inventions in each case and the manner in which the inventions were recited in the challenged claims are the factors that drove their different outcomes. Practitioners should appreciate this takeaway at all stages of patent preparation and prosecution. Careful scrutiny is needed in both assessing what is patentable and how to claim it to demonstrate its patentability. As RecogniCorp illustrates, the current state of the law on patent-eligibility eschews categorical rules and instead is driven by case-specific nuances.