Google LLC v. Human Differential Intelligence LLC

1. Case Identification

• Case #: IPR2020-01627
• Patent #: 8,752,141
• Filed: September 15, 2020
• Petitioner(s): Google LLC
• Patent Owner(s): Human Differential Intelligence, LLC
• Challenged Claims: 1-5, 7-14

2. Patent Overview

• Title: Image-Based Progressive CAPTCHA Efficacy Testing
• Brief Description: The ’141 patent relates to methods for presenting and determining the efficacy of an image-based CAPTCHA (Completely Automatic Public Turing Test To Tell Humans And Computers Apart). The system progressively reveals an image to a user and measures effectiveness by calculating the percentage of users who can correctly identify an associated concept within a predetermined time.

3. Grounds for Unpatentability

Ground 1: Obviousness over Qvarfordt - Claims 1-5 and 7-14 are obvious over Qvarfordt.

• Prior Art Relied Upon: Qvarfordt (Application # 2008/0127302).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Qvarfordt disclosed all key limitations of the challenged claims. Qvarfordt taught an animated, image-based CAPTCHA system that progressively reveals an "encoded solution" (e.g., an image of a cat) by moving obstructing foreground elements. This action of partially obscuring the image met the claim limitation of presenting a "subset" of imagery data that is "progressively revealed." Petitioner asserted Qvarfordt's two-set system—using "novel" CAPTCHAs to gather user responses and then moving them to a "vetted" set for access control—disclosed the claimed method of "testing recognition of images." In this process, Qvarfordt associated a "solution" or "solution set" (e.g., "cat" or "kitten") with the image, meeting the limitation of associating the image with a concept.
  • Key Aspects: For the final limitation of "calculating a percentage of participants...within a predetermined time," Petitioner contended this was an obvious implementation detail. A person of ordinary skill in the art (POSA) would have understood that to achieve Qvarfordt's goal of ensuring a high human success rate (e.g., 90%), it would be necessary to measure that success rate. Calculating a percentage was a conventional and trivial way to do so. Similarly, a POSA would have imposed a time limit for responses as a standard security measure to prevent bot attacks, making this feature obvious to add to Qvarfordt's system.

Ground 2: Obviousness over Qvarfordt in view of vonAhn - Claims 1-5 and 7-14 are obvious over Qvarfordt in view of vonAhn.

• Prior Art Relied Upon: Qvarfordt (Application # 2008/0127302) and vonAhn (a 2003 EUROCRYPT publication).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Qvarfordt taught the foundational CAPTCHA system, as detailed in Ground 1. To the extent Qvarfordt was found not to explicitly teach calculating a success percentage within a predetermined time, vonAhn supplied these elements. vonAhn was cited for its explicit teaching on the importance of empirically verifying CAPTCHA efficacy by measuring the percentage of the human population that can pass the test. Further, vonAhn disclosed the "PIX" CAPTCHA, which required users to respond "within a short time after they are presented" and before a timer expired to be deemed human.
  • Motivation to Combine: A POSA would combine vonAhn's teachings with Qvarfordt's system for two primary reasons. First, to ensure Qvarfordt’s system met its own stated goal of being "as easy as possible" for humans, a POSA would be motivated by vonAhn to empirically measure the success rate. Second, to enhance security—a primary goal of any CAPTCHA system, including Qvarfordt's—a POSA would incorporate vonAhn's use of a timer to prevent bots from using unlimited time to solve the challenge.
  • Expectation of Success: A POSA would have a reasonable expectation of success in this combination. Implementing a simple calculation (dividing successful users by total users) and adding a standard software timer were well-known techniques that would predictably improve the efficacy and security of Qvarfordt's image-based CAPTCHA system.

4. Arguments Regarding Discretionary Denial

• §314(a) Fintiv Factors: Petitioner argued that discretionary denial under Fintiv would be inappropriate. Key reasons included that Google is not a defendant in the parallel district court litigation, the petition was filed expeditiously (seven months before any potential statutory bar), the district court cases were in their infancy with minimal investment, and the trial date was speculative and likely to be delayed.
• §325(d) Factors: Petitioner asserted that denial under §325(d) was unwarranted because the petition raised materially new arguments. The primary reference, Qvarfordt, was merely cited in an Information Disclosure Statement (IDS) during prosecution and was never applied or substantively discussed by the Examiner. The secondary reference, vonAhn, was never before the Examiner. Petitioner argued that because the Examiner never considered the teachings of these references, the office did not have a previous opportunity to evaluate the invalidity arguments presented in the petition.

5. Relief Requested

• Petitioner requests institution of an inter partes review (IPR) and cancellation of claims 1-5, 7-14 of Patent 8,752,141 as unpatentable.