accessiBe Ltd v. AudioEye Inc

1. Case Identification

• Case #: IPR2022-01432
• Patent #: 10,997,361
• Filed: September 6, 2022
• Petitioner(s): accessiBe Ltd.
• Patent Owner(s): AudioEye, Inc.
• Challenged Claims: 1-20

2. Patent Overview

• Title: Modular Systems and Methods for Selectively Enabling Cloud-Based Assistive Technologies
• Brief Description: The ’361 patent relates to systems and methods for automatically remediating websites to improve accessibility for individuals with disabilities. The technology involves detecting accessibility compliance issues, such as untagged images, and programmatically assigning Accessible Rich Internet Applications (ARIA) labels using image recognition algorithms to provide audible descriptions.

3. Grounds for Unpatentability

Ground 1: Obviousness of Claims 1, 3-6, and 8-19 over Springer, Hendry, and Folkens

• Prior Art Relied Upon: Springer (Application # 2004/0148568), Hendry (Application # 2013/0104029), and Folkens (Application # 2015/0286658).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Springer taught a foundational system of "checkers" and "fixers" for identifying and remediating web accessibility issues, such as images lacking alternative (alt) text, by modifying a webpage’s Document Object Model (DOM). However, Springer’s methods were dated. Hendry was cited for disclosing more modern techniques, including the use of ARIA labels (the specific type recited in the claims) and a distributed computing architecture where remediation tasks could be performed on remote servers. Folkens was introduced for its disclosure of an artificial intelligence (AI)-based image recognition system that automatically generates descriptive tags for images. Petitioner contended that combining these references taught all limitations of independent claim 1, including accessing web page code (Springer), detecting untagged images (Springer), determining an image’s subject matter with an AI algorithm (Folkens), and assigning a descriptive ARIA label to alter the code for use by assistive technologies (Hendry).
  • Motivation to Combine: A Person of Ordinary Skill in the Art (POSITA) would combine these references to modernize and improve the functionality of Springer’s accessibility tool. A POSITA would integrate Folkens' AI capabilities to automate the creation of descriptive text for images, a known limitation of systems like Springer's that often required manual input. Further, a POSITA would incorporate Hendry’s teachings on ARIA labels, a more modern standard than simple alt-text, and would utilize Hendry’s distributed computing model to offload computationally intensive AI processing to a remote server, improving efficiency.
  • Expectation of Success: A POSITA would have a reasonable expectation of success in combining these known elements, as it involved applying predictable solutions (AI image analysis, remote processing) to improve a known system (web accessibility remediation).

Ground 2: Obviousness of Claims 2, 7, and 20 over Springer, Folkens, Hendry, and Beeman

• Prior Art Relied Upon: Springer (Application # 2004/0148568), Folkens (Application # 2015/0286658), Hendry (Application # 2013/0104029), and Beeman (Application # 2009/0319927).

• Core Argument for this Ground:

  • Prior Art Mapping: This ground built upon the core combination of Springer, Folkens, and Hendry to address dependent claims requiring the method to be performed on a "periodic basis" (claims 2 and 20) and for the image recognition to be based on "contextual cues" (claim 7). Petitioner asserted that Beeman disclosed a document checking system that specifically taught performing an "asynchronous scan of a document for errors" on a periodic basis. Beeman also taught locating the "context" for each identified error to provide information on how to fix it. Petitioner argued that adding Beeman’s teachings to the primary combination rendered these dependent claims obvious.
  • Motivation to Combine: A POSITA would be motivated to add Beeman’s teachings to the core combination to further enhance the system’s utility. Incorporating periodic scanning would allow for continuous monitoring of a website’s accessibility, which is more effective than a one-time scan, especially for frequently updated sites. Furthermore, using contextual information from the webpage, as taught by Beeman and Hendry, would provide valuable input to Folkens' AI algorithm, improving the accuracy of its image recognition and the quality of the generated descriptions.
  • Expectation of Success: A POSITA would expect success in integrating a periodic scanning function into an accessibility tool, as this represented a straightforward and predictable improvement to ensure ongoing compliance.

• Additional Grounds: Petitioner asserted additional obviousness challenges (Grounds 2 and 4) that relied on similar arguments but substituted Lehota (Application # 2010/0205523) for Springer as the primary reference teaching a system that modifies a website's DOM to improve accessibility via embedded widgets.

4. Arguments Regarding Discretionary Denial

• Petitioner argued that discretionary denial of institution under 35 U.S.C. §314(a) based on the Fintiv factors would be inappropriate. Petitioner contended that the trial date in the parallel district court litigation was uncertain, and the case was likely to be transferred, weighing against denial. It was also argued that there had been no significant investment in the district court case, as no discovery had occurred and the claim construction hearing had been cancelled. Finally, Petitioner asserted that it presented a strong case for unpatentability and that the specific arguments and combinations were not yet raised in the district court, thus favoring institution of the inter partes review (IPR).

5. Relief Requested

• Petitioner requests institution of an IPR and cancellation of claims 1-20 of the ’361 patent as unpatentable.