DCT

1:26-cv-00868

VB Assets LLC v. IBM Corp

Log In
Key Events
Complaint
complaint Intelligence

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: 1:26-cv-00868, D. Del., 07/16/2026
  • Venue Allegations: Venue is alleged to be proper in the District of Delaware because Defendant IBM has committed acts of infringement in the district, offers the accused products for sale there, and maintains regular and established places of business in Wilmington and Newark, Delaware. The complaint also notes that IBM has not contested venue in the district in prior litigation.
  • Core Dispute: Plaintiff alleges that Defendant's Watson-branded voice assistant and artificial intelligence products and services infringe eight U.S. patents related to cooperative conversational voice user interfaces, natural language understanding, and voice-based advertisement delivery.
  • Technical Context: The technology at issue concerns natural language understanding (NLU) for voice-based AI systems, which enables computers to interpret and respond to human speech in a conversational manner, a foundational technology for modern digital assistants.
  • Key Procedural History: The complaint alleges a long history between the parties, beginning with a collaborative partnership between Plaintiff's predecessor (VoiceBox Technologies) and IBM starting as early as 2003. It further alleges that IBM was aware of the patent portfolio, citing the patents-in-suit as prior art during its own patent prosecution. Plaintiff also notes it previously asserted some of the same patents against Amazon.com, Inc. in the same district. An inter partes review of the U.S. Patent 8,073,681 resulted in the disclaimer of claims 37-42, with claims 1-36 confirmed as patentable.

Case Timeline

Date Event
2003-01-01 Alleged collaboration between IBM and VoiceBox Technologies begins.
2006-10-16 Earliest Priority Date for '681, '699, '341, '628 Patents.
2007-02-06 Earliest Priority Date for '176, '097, '489, '274 Patents.
2010-10-19 U.S. Patent 7,818,176 Issues.
2011-12-06 U.S. Patent 8,073,681 Issues.
2012-03-27 U.S. Patent 8,145,489 Issues.
2013-01-01 IEEE ranks VoiceBox Technologies #13 in patent power for computer software.
2013-09-03 U.S. Patent 8,527,274 Issues.
2016-02-23 U.S. Patent 9,269,097 Issues.
2019-07-29 Plaintiff VB Assets, LLC files suit against Amazon.com, Inc.
2019-12-17 U.S. Patent 10,510,341 Issues.
2019-12-24 U.S. Patent 10,515,628 Issues.
2020-07-28 IPR trial (IPR2020-01367) instituted for the '681 Patent.
2020-08-25 U.S. Patent 10,755,699 Issues.
2022-12-21 Inter Partes Review Certificate for '681 Patent issues, disclaiming claims 37-42.
2026-07-16 Complaint Filing Date.

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 8,073,681 - "System and Method for a Cooperative Conversational Voice User Interface"

  • Patent Identification: U.S. Patent No. 8,073,681, titled "System and Method for a Cooperative Conversational Voice User Interface," issued December 6, 2011 Compl. ¶16

The Invention Explained

  • Problem Addressed: The patent's background describes prior speech interfaces as rigid, "Command and Control" systems that forced users to memorize exact phrases and navigate menus, failing to provide the intuitive, cooperative dialogue characteristic of human conversation Compl. ¶¶28-29 '681 Patent, col. 1:44-60
  • The Patented Solution: The invention is a "cooperative conversational" voice interface that uses both short-term shared knowledge (from the current conversation) and long-term shared knowledge (from past conversations) to understand a user's intent Compl. ¶17 '681 Patent, abstract By identifying a context for an utterance, the system can disambiguate words with multiple meanings and generate an adapted, conversational response, a process illustrated in the patent's figures detailing the cooperative conversation model '681 Patent, Fig. 3
  • Technical Importance: This technology represented a shift from simple voice command execution to a more nuanced, context-aware conversational paradigm, which is a foundational concept for modern voice assistants Compl. ¶¶30-32

Key Claims at a Glance

  • The complaint alleges infringement of patented methods and systems, with infringement counts pointing to claim chart exhibits that are not provided Compl. ¶¶65-66 The narrative description of the invention tracks the elements of independent claim 1 Compl. ¶17
  • Essential elements of independent claim 1 include:
    • Receiving an utterance with words having different meanings in different contexts.
    • Accumulating short-term shared knowledge about the current conversation.
    • Accumulating long-term shared knowledge about the user's past conversations.
    • Identifying a context for the utterance using both short-term and long-term knowledge.
    • Establishing an intended meaning for the utterance within that context to disambiguate intent.
    • Generating a response that is grammatically or syntactically adapted based on the intended meaning.
  • The complaint reserves the right to assert additional claims (Compl. ¶65).

U.S. Patent No. 10,755,699 - "System and Method for a Cooperative Conversational Voice User Interface"

  • Patent Identification: U.S. Patent No. 10755699, titled "System and Method for a Cooperative Conversational Voice User Interface," issued August 25, 2020 Compl. ¶19

The Invention Explained

  • Problem Addressed: As with the '681 Patent, the '699 Patent addresses the shortcomings of prior art voice interfaces that were not intuitive and required users to "dumb down" their requests to fit rigid command structures Compl. ¶¶28-29
  • The Patented Solution: The '699 Patent claims a method for generating responses adapted to a user's manner of speaking Compl. ¶20 In addition to using short-term and long-term knowledge to determine an interpretation, the system specifically identifies the "manner in which the natural language utterance was spoken" and generates a response based on both the interpretation and this identified manner '699 Patent, col. 6:32-37 The patent suggests this can include varying tone, pace, or inflection in the response '699 Patent, col. 6:35-37
  • Technical Importance: This invention adds a layer of sophistication by responding not just to what was said, but how it was said, aiming to make human-machine interaction feel more natural and personalized (Compl. ¶31; Compl. ¶32).

Key Claims at a Glance

  • The complaint alleges infringement of methods claimed in the '699 Patent, with the narrative description tracking independent claim 1 (Compl. ¶¶20; Compl. ¶74).
  • Essential elements of independent claim 1 include:
    • Receiving and recognizing words from a natural language utterance.
    • Identifying a context and determining an interpretation for the utterance.
    • Accumulating short-term (single conversation) and long-term (prior conversations) knowledge.
    • Identifying a "manner in which the natural language utterance was spoken" based on the short-term and long-term knowledge.
    • Generating a response based on both the interpretation and the identified manner of speaking.
  • The complaint reserves the right to assert additional claims Compl. ¶74

U.S. Patent No. 10,510,341 - "System and Method for a Cooperative Conversational Voice User Interface"

  • Patent Identification: U.S. Patent No. 10510341, titled "System and Method for a Cooperative Conversational Voice User Interface," issued December 17, 2019 Compl. ¶22
  • Technology Synopsis: The '341 Patent claims a system for facilitating natural language responses that uses accumulated short-term and long-term knowledge Compl. ¶23 A key aspect is the management of this knowledge, wherein the system is configured to "expire" items of short-term knowledge and include those expired items in the accumulated long-term knowledge, which is then used to determine the context and interpretation of a new utterance '341 Patent, abstract Compl. ¶23
  • Asserted Claims: The complaint describes a system that embodies the features of independent claim 10 Compl. ¶23 Compl. ¶83
  • Accused Features: The complaint alleges that IBM's Watson products accumulate, expire, and use short-term and long-term knowledge to process user utterances (Compl. ¶¶5; Compl. ¶83; Compl. fn. 5).

U.S. Patent No. 10,515,628 - "System and Method for a Cooperative Conversational Voice User Interface"

  • Patent Identification: U.S. Patent No. 10515628, titled "System and Method for a Cooperative Conversational Voice User Interface," issued December 24, 2019 Compl. ¶25
  • Technology Synopsis: The '628 Patent is similar to the '341 Patent in its use of expiring short-term knowledge that becomes part of long-term knowledge Compl. ¶26 It distinguishes itself by claiming a system that accumulates and uses "first long-term knowledge associated with a first user" and "second long-term knowledge associated with a second user" to determine the context for an utterance, suggesting a multi-user awareness '628 Patent, abstract Compl. ¶26
  • Asserted Claims: The complaint describes a system that embodies the features of independent claim 1 Compl. ¶26 Compl. ¶92
  • Accused Features: The infringement allegations target the multi-user capabilities of IBM's Watson platform, which allegedly utilize user-specific knowledge in conversational contexts (Compl. ¶¶5; Compl. ¶92; Compl. fn. 6).

U.S. Patent No. 7,818,176 - "System and Method for Selecting and Presenting Advertisements Based on Natural Language Processing of Voice-Based Input"

  • Patent Identification: U.S. Patent No. 7818176, titled "System and Method for Selecting and Presenting Advertisements Based on Natural Language Processing of Voice-Based Input," issued October 19, 2010 Compl. ¶34
  • Technology Synopsis: The '176 Patent claims a system for delivering advertisements in response to voice-based natural language requests Compl. ¶35 The system uses a speech recognition engine that maps phonemes to syllables to generate a preliminary interpretation, and a conversational language processor then establishes a context to select and present a relevant advertisement '176 Patent, abstract Compl. ¶35
  • Asserted Claims: The complaint's description of the invention follows the elements of independent claims 1 and 27 Compl. ¶35 Compl. ¶36
  • Accused Features: The allegations target IBM's use of its Watson platform to deliver "Cognitive Ads" or other targeted advertisements in response to voice queries, such as in the automotive context Compl. ¶103 Compl. fn. 8

U.S. Patent No. 9,269,097 - "System and Method for Delivering Targeted Advertisements and/or Providing Natural Language Processing Based on Advertisements"

  • Patent Identification: U.S. Patent No. 9269097, titled "System and Method for Delivering Targeted Advertisements and/or Providing Natural Language Processing Based on Advertisements," issued February 23, 2016 Compl. ¶38
  • Technology Synopsis: The '097 Patent claims a method for providing natural language processing based on a presented advertisement, specifically by interpreting a subsequent user utterance that contains a pronoun Compl. ¶39 The system determines whether the pronoun refers to the product, service, or provider from the advertisement, addressing the ambiguity of pronouns in a conversational context ('097 Patent, abstract; Compl. ¶¶39; Compl. ¶52).
  • Asserted Claims: The complaint's description follows independent claim 1 Compl. ¶39 Compl. ¶110
  • Accused Features: The allegations focus on IBM's advertising systems that allow for conversational follow-up questions to ads, where the system must resolve pronoun references Compl. ¶112 Compl. fns. 8-9

U.S. Patent No. 8,145,489 - "System and Method for Selecting and Presenting Advertisements Based on Natural Language Processing of Voice-Based Input"

  • Patent Identification: U.S. Patent No. 8145489, titled "System and Method for Selecting and Presenting Advertisements Based on Natural Language Processing of Voice-Based Input," issued March 27, 2012 Compl. ¶41
  • Technology Synopsis: The '489 Patent claims a server-based system that delivers an advertisement to an electronic device, tracks the user's interaction pattern with that ad, and updates various models based on that interaction Compl. ¶42 These updated models-which include information about a specific user, multiple users, and environmental conditions-are then used to interpret subsequent utterances '489 Patent, abstract Compl. ¶42
  • Asserted Claims: The complaint's description follows independent claim 9 Compl. ¶42 Compl. ¶119
  • Accused Features: The allegations target IBM's advertising platforms that track user engagement with ads to update user profiles and other models for future ad targeting and query interpretation Compl. ¶121 Compl. fn. 8

U.S. Patent No. 8,527,274 - "System and Method for Delivering Targeted Advertisements and Tracking Advertisement Interactions in Voice Recognition Contexts"

  • Patent Identification: U.S. Patent No. 8527274, titled "System and Method for Delivering Targeted Advertisements and Tracking Advertisement Interactions in Voice Recognition Contexts," issued September 3, 2013 Compl. ¶44
  • Technology Synopsis: The '274 Patent claims a method for handling incomplete or ambiguous user requests by presenting one or more associated advertisements Compl. ¶45 The system then monitors the user's interaction with the ads and interprets the original ambiguous request based on that interaction, using the ad interaction as a form of clarification ('274 Patent, abstract; Compl. ¶¶45; Compl. ¶54).
  • Asserted Claims: The complaint's description follows independent claim 1 Compl. ¶45 Compl. ¶128
  • Accused Features: The allegations target IBM's systems that present advertisements as a way to clarify or respond to ambiguous user queries and then use the subsequent user interaction to refine the interpretation of the initial request Compl. ¶130 Compl. fn. 8

III. The Accused Instrumentality

Product Identification

  • The accused instrumentalities are Defendant IBM's Watson-branded products and services, including "Watson Assistant, watsonx Assistant, Watson Assistant for Voice Interaction, watsonx Orchestrate, Voice Agent with Watson," and the underlying software, hardware, servers, and computer-readable instructions that implement them Compl. ¶5

Functionality and Market Context

  • The Accused IBM Products are described as voice assistants, voice-based AI systems, and voice-recognition technologies that enable users to interact with computer services through natural language speech Compl. ¶5 The complaint alleges their functionality includes using "short-term and/or long-term shared knowledge" to determine context, infer information, and provide adaptive responses Compl. ¶31 This is supported by citations to IBM's own technical documentation describing, for example, the use of "context variables" to maintain state during a conversation and personalize the dialogue Compl. fn. 2, p. 26 Compl. fn. 4, p. 28 The complaint's Figure 1, which shows an early "Cybermind" prototype, is presented as evidence of the plaintiff's pioneering work on the core concepts now allegedly embodied in these modern systems Compl. ¶13, p. 4
  • The complaint positions the Accused IBM Products as "increasingly popular and well-known" and fundamental to "modern-day voice assistants" Compl. ¶4 The complaint also references an IEEE ranking that placed Plaintiff's predecessor, VoiceBox Technologies, at number 13 for patent power in the computer software industry in 2013, suggesting the commercial and technical significance of the underlying patented technology Compl. ¶14, p. 5

IV. Analysis of Infringement Allegations

U.S. Patent 8,073,681 Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
...receiving an utterance at a voice input device, during a current conversation...wherein the utterance includes one or more words that have different meanings in different contexts... The Accused IBM Products receive voice utterances from users as part of a conversational flow. ¶65; ¶66 col. 2:35-39
...accumulating short-term shared knowledge about the current conversation... The Accused IBM Products allegedly maintain and update "context variables" during a single conversational session to store information that can be used by the dialog. ¶65; ¶67; fn. 2 col. 5:15-24
...accumulating long-term shared knowledge about the user, wherein the long-term shared knowledge includes knowledge about one or more past conversations with the user... The Accused IBM Products allegedly build and maintain user profiles and historical data over time, which are used to personalize and inform conversations. ¶65; ¶67; fn. 2 col. 5:35-43
...identifying a context associated with the utterance...from the short-term shared knowledge and the long-term shared knowledge... The Accused IBM Products allegedly use session-level context variables and long-term user data to determine the context and control the conversational flow. ¶65; ¶67; fn. 2 col. 4:5-14
...establishing an intended meaning for the utterance within the identified context...to disambiguate an intent... The Accused IBM Products allegedly use intent detection and context to interpret the user's goal, even for ambiguous utterances. ¶65; ¶67; fn. 2 col. 4:15-24
...generating a response to the utterance, wherein the conversational speech engine grammatically or syntactically adapts the response based on the intended meaning... The Accused IBM Products generate responses based on the determined intent and context. ¶65; ¶67; fn. 2 col. 6:25-30

U.S. Patent 10,755,699 Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
...receiving...a user input comprising a natural language utterance; recognizing...one or more words or phrases from the natural language utterance... The Accused IBM Products receive and perform speech-to-text on user utterances. ¶74; ¶75; fn. 4 col. 1:45-52
...identifying...a context for the natural language utterance...; determining...an interpretation...based on the identified context... The Accused IBM Products use context variables and intent detection to interpret the user's utterance. ¶74; ¶75; fn. 4 col. 4:5-24
...accumulating...short-term knowledge based on one or more natural language utterances received during a predetermined time period...; accumulating...long-term knowledge... The Accused IBM Products allegedly use session-level context variables (short-term) and persistent user profiles (long-term). ¶74; ¶75; fn. 4 col. 5:15-43
...identifying...a manner in which the natural language utterance was spoken based on the short-term knowledge and the long-term knowledge... The Accused IBM Products allegedly include a "Tone Analyzer" and "Tone Classification" functionality to analyze user input beyond the literal words. ¶74; ¶76; fn. 4 col. 6:32-37
...generating...a response to the natural language utterance based on the interpretation and the identified manner in which the natural language utterance was spoken. The Accused IBM Products generate responses based on the interpretation of the user's intent and, allegedly, the manner of speaking. ¶74; ¶75; fn. 4 col. 6:25-30
  • Identified Points of Contention:
    • Scope and Equivalence Questions: A primary area of dispute may involve the precise definitions of "accumulating short-term shared knowledge" and "long-term shared knowledge." The infringement analysis will likely question whether the "context variables" and user profiles in IBM's systems Compl. fn. 2, p. 26 function in the specific manner claimed by the patents, particularly with respect to how knowledge is built, maintained, and expired (as claimed in the '341 and '628 Patents).
    • Technical and Evidentiary Questions: For the '699 Patent, a key question will be evidentiary: what proof demonstrates that the accused Watson products actually use the "manner in which the utterance was spoken" to generate a response, as required by the claim? While the complaint points to IBM's "Tone Analyzer" technology Compl. fn. 4, p. 29, the analysis will turn on whether this feature is integrated into the accused conversational flows in the specific way the patent claims, or if it is a separate, non-integrated analytical tool.

V. Key Claim Terms for Construction

  • The Term: "accumulating short-term shared knowledge" and "accumulating long-term shared knowledge" ('681 Patent, Claim 1)
  • Context and Importance: These terms are the foundation of the '681 patent family's claimed invention, distinguishing it from prior art systems that allegedly treated each utterance in isolation (Compl. ¶32). The outcome of the case may depend on whether IBM's method of using session-specific "context variables" and persistent user profiles falls within the scope of these terms as defined by the patents.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification describes shared knowledge as including information that can "continually build and draw upon shared information," which could be argued to encompass any system that maintains conversational state and user history '681 Patent, col. 1:58-61
    • Evidence for a Narrower Interpretation: The specification details that short-term data may be "expired after a psychologically appropriate amount of time" and that long-term models are "user-centric, rather than session-based" '681 Patent, col. 5:30-43 A defendant may argue these passages imply a specific architectural relationship between expiring short-term data and building long-term models that is not present in the accused systems.
  • The Term: "manner in which the natural language utterance was spoken" ('699 Patent, Claim 1)
  • Context and Importance: This term is the central inventive concept of the '699 Patent. Practitioners may focus on this term because it distinguishes this patent from others in the portfolio that focus only on the semantic content of an utterance. The complaint's specific allegation that IBM's "Tone Analyzer" infringes suggests this term will be a focal point of the dispute Compl. fn. 4, p. 29
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The term could be argued to cover any non-literal aspect of the utterance that influences the system's response, potentially including syntax and word choice as a "manner" of speaking.
    • Evidence for a Narrower Interpretation: The specification provides specific examples of what this "manner" entails, stating that intelligent responses may have "natural variation and/or personality (e.g., by varying tone, pace, timing, inflection, word use, jargon, and other variables...)" '699 Patent, col. 6:35-37 This language may support a narrower construction limited to measurable acoustic and paralinguistic features.

VI. Other Allegations

  • Indirect Infringement: The complaint alleges that IBM induces infringement by actively encouraging its customers and licensees to use the Accused IBM Products in an infringing manner Compl. ¶67 Compl. ¶76 This encouragement is allegedly provided through "technical support, documentation, and consulting and support services" Compl. ¶67 The complaint also asserts a theory of divided infringement, alleging IBM directs or controls the performance of method steps by end users by designing the software to require user input to function (Compl. ¶66).
  • Willful Infringement: Willfulness is extensively alleged based on both pre- and post-suit knowledge. The complaint alleges pre-suit knowledge stemming from a long-standing business and collaborative relationship between IBM and Plaintiff's predecessor, VoiceBox Technologies, dating back to 2003 Compl. ¶¶57-59 Further, it is alleged that IBM cited the patents-in-suit as prior art during the prosecution of its own patents, demonstrating direct knowledge of the portfolio Compl. ¶61 The complaint also pleads knowledge based on industry awareness of Plaintiff's prior litigation against Amazon Compl. ¶62, and actual notice as of the complaint's filing date Compl. ¶69

VII. Analyst's Conclusion: Key Questions for the Case

  • A core issue will be one of technical implementation and claim scope: Does the architecture of IBM's Watson platform, particularly its use of "context variables" and user profiles, meet the specific claim limitations for "accumulating short-term and long-term shared knowledge," including the knowledge management and expiration cycles described in patents like the '341 Patent? For the '699 patent, this translates to an evidentiary question: does the complaint show that IBM's "Tone Analyzer" is not just an available feature, but is integrated into the accused conversational products to adapt responses based on the "manner" of speaking, as the claim requires?
  • A second central question will be one of willfulness and intent: Given the extensive allegations of a decade-long prior relationship, public partnerships, and IBM's own citations to the patents-in-suit during its patent prosecution, a key issue for the court will be determining whether IBM acted with knowledge of, or willful blindness to, its alleged infringement, which could significantly impact potential damages.
  • For the patents related to advertising, a key question will be one of functional specificity: Do IBM's "Cognitive Ads" or other voice-based advertising systems practice the specific claimed methods of using context, resolving pronouns, and interpreting ambiguous requests based on user interactions with advertisements, or do they operate on a more general principle of contextual advertising that falls outside the patent claims?
Loading Complaint