VB Assets LLC v. Google LLC

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: VB Assets, LLC (Delaware)
  • Defendant: Google LLC; and YouTube, LLC (Delaware)
  • Plaintiff's Counsel: Farnan LLP
• Case Identification: 1:26-cv-00443, D. Del., 07/24/2026
• Venue Allegations: Venue is asserted on the basis that both Defendants are incorporated and reside in the District of Delaware.
• Core Dispute: Plaintiff alleges that Defendant's AI products, including Google Assistant, Gemini, and Google/YouTube voice search, infringe a portfolio of eleven patents related to conversational AI, voice-based advertising, and voice commerce.
• Technical Context: The technology at issue is conversational artificial intelligence, which enables more natural, human-like interaction with voice-activated devices and services, a fundamental component of the modern digital assistant market.
• Key Procedural History: The complaint alleges an extensive pre-suit history, including meetings and technology demonstrations between Plaintiff's predecessor, VoiceBox Technologies, and Google executives beginning in 2007. Plaintiff asserts that it repeatedly disclosed its patent portfolio to Google, starting with pending applications in 2010. The complaint also alleges that Google has cited the patents-in-suit as prior art on numerous occasions during the prosecution of its own patents, which may be significant for allegations of knowledge and willfulness.

Case Timeline

Date	Event
2006-10-16	Priority Date for '681, '765, '249, '341, '699 Patents
2007-05-01	Priority Date for '176, '274, '536, '097, '456 Patents
2007-01-01	VoiceBox Technologies allegedly first met with Google's Head of Speech
2010-10-19	'176 Patent Issued
2011-01-01	Google allegedly announced Google voice search
2011-12-06	'681 Patent Issued
2013-08-20	'765 Patent Issued
2013-09-03	'274 Patent Issued
2014-11-11	'536 Patent Issued
2015-01-12	Priority Date for '385 Patent
2016-01-01	Google Assistant launched
2016-02-23	'097 Patent Issued
2017-08-01	Google announced partnership with Walmart for voice shopping
2018-03-02	VoiceBox allegedly sent notice of patent portfolio to Google
2019-05-21	'249 Patent Issued
2019-12-17	'341 Patent Issued
2020-08-25	'699 Patent Issued
2021-08-10	'385 Patent Issued
2023-02-01	Google announced Bard
2024-02-01	Google rebranded Bard as Google Gemini
2025-02-25	'456 Patent Issued
2026-07-24	First Amended Complaint Filed

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

• Issued: December 6, 2011 (the "'681 Patent")

The Invention Explained

• Problem Addressed: The patent's background section, as described in the complaint, addresses the limitations of typical computer speech-recognition systems that force users to use a limited number of rigid prompts, inhibiting widespread adoption Compl. ¶14
• The Patented Solution: The invention proposes a "cooperative conversational voice user interface" that determines user intent by leveraging both "short-term" and "long-term" knowledge Compl. ¶24 This allows the system to understand natural, free-form utterances by identifying the context of the conversation. The system uses this context to disambiguate words with multiple meanings and generate a grammatically adapted response Compl. ¶24 '249 Patent, col. 5:4-9 This process is designed to model a more natural, human-like dialogue Compl. ¶14
• Technical Importance: This approach enabled a move away from rigid command-and-control voice systems toward more flexible, context-aware conversational interfaces Compl. ¶14

Key Claims at a Glance

• The complaint asserts independent claim 1 Compl. ¶24 Compl. ¶95
• Essential elements of claim 1 include:
  • receiving an utterance during a current conversation that includes words with 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;
  • determining an intended meaning by identifying a context from the short-term and long-term shared knowledge;
  • establishing the intended meaning within that context to disambiguate user intent; and
  • generating a response that is grammatically or syntactically adapted based on the established meaning.

U.S. Patent No. 8,515,765 - SYSTEM AND METHOD FOR A COOPERATIVE CONVERSATIONAL VOICE USER INTERFACE

• Issued: August 20, 2013 (the "'765 Patent")

The Invention Explained

• Problem Addressed: Similar to the '681 Patent, the invention seeks to make human-computer voice interactions more natural and effective Compl. ¶14 This patent focuses specifically on managing the turn-taking and flow of the conversation.
• The Patented Solution: The patented method determines a "conversation type" by assessing whether the user has a "leader role" or a "supporter role" in the interaction Compl. ¶27 Based on this role, the system generates a response with a specific format. If the user has a "supporter role," the response format is adapted to limit their future input to "interjecting queries or requests for clarification," thereby guiding the conversation more efficiently '249 Patent, col. 15:35-50 Compl. ¶27
• Technical Importance: This system provides a framework for dynamically managing conversational flow, allowing the system to take the lead when appropriate, which can streamline interactions and guide the user to a resolution.

Key Claims at a Glance

• The complaint asserts independent claim 1 Compl. ¶27 Compl. ¶111
• Essential elements of claim 1 include:
  • receiving a natural language utterance corresponding to a conversation type;
  • determining the conversation type based on whether a user has a "leader role" or a "supporter role" in the interaction; and
  • generating a response with a format based on the conversation type, where the format is adapted to limit the user's future input if the user has the supporter role.

U.S. Patent No. 10,297,249 - SYSTEM AND METHOD FOR A COOPERATIVE CONVERSATIONAL VOICE USER INTERFACE

• Issued: May 21, 2019 (the "'249 Patent")
• Technology Synopsis: The patent describes a method for a system to facilitate natural language responses by using short-term knowledge generated from prior multi-modal (e.g., voice and non-voice) interactions. The method involves filtering sound from multiple voice inputs and using the state of a user interface to determine context and generate a response Compl. ¶30
• Asserted Claims: Claim 1 is asserted Compl. ¶30 Compl. ¶127
• Accused Features: The complaint alleges that Google AI Products, such as Google Assistant, which can respond to multi-modal inputs, infringe the '249 Patent Compl. ¶131 Compl. ¶133

U.S. Patent No. 10,510,341 - SYSTEM AND METHOD FOR A COOPERATIVE CONVERSATIONAL VOICE USER INTERFACE

• Issued: December 17, 2019 (the "'341 Patent")
• Technology Synopsis: This patent claims a method for using both accumulated short-term and long-term knowledge to generate natural language responses. A key aspect is the "expiring" of older short-term knowledge, which can then be incorporated into the long-term knowledge base, mimicking a form of memory decay and learning Compl. ¶33
• Asserted Claims: Claim 1 is asserted Compl. ¶33 Compl. ¶142
• Accused Features: Google Assistant's "Continued Conversation" and "Personal References" features, which accumulate and use both short-term and long-term knowledge, are accused of infringement Compl. ¶146 Compl. ¶147

U.S. Patent No. 10,755,699 - SYSTEM AND METHOD FOR A COOPERATIVE CONVERSATIONAL VOICE USER INTERFACE

• Issued: August 25, 2020 (the "'699 Patent")
• Technology Synopsis: The patent describes a method of generating responses that are adapted based on a user's "manner of speaking." The system uses short-term and long-term knowledge to identify how an utterance was spoken and generates a response based on both the interpretation and the identified manner Compl. ¶36
• Asserted Claims: Claim 1 is asserted Compl. ¶36 Compl. ¶157
• Accused Features: The complaint alleges Google AI Products, which use contextual information to understand and respond to utterances, infringe this patent Compl. ¶161 Compl. ¶163

U.S. Patent No. 7,818,176 - SYSTEM AND METHOD FOR SELECTING AND PRESENTING ADVERTISEMENTS BASED ON NATURAL LANGUAGE PROCESSING OF VOICE-BASED INPUT

• Issued: October 19, 2010 (the "'176 Patent")
• Technology Synopsis: This patent relates to "Voice Ads," claiming a method for presenting advertisements in response to natural language voice queries. The method involves recognizing words from an utterance, establishing a context, and then selecting and presenting an advertisement relevant to that context Compl. ¶39
• Asserted Claims: Claim 1 is asserted Compl. ¶39 Compl. ¶172
• Accused Features: Google's features that provide sponsored results and targeted advertisements in response to voice queries in products like Google Assistant and Google Search are accused of infringement Compl. ¶177 Compl. ¶178

U.S. Patent No. 8,527,274 - SYSTEM AND METHOD FOR DELIVERING TARGETED ADVERTISEMENTS AND TRACKING ADVERTISEMENT INTERACTIONS IN VOICE RECOGNITION CONTEXTS

• Issued: September 3, 2013 (the "'274 Patent")
• Technology Synopsis: The patent describes a method for delivering targeted ads when a voice request is incomplete or ambiguous. The system presents advertisements, monitors user interaction with those ads, and then uses that interaction to interpret the original ambiguous request Compl. ¶42
• Asserted Claims: Claim 1 is asserted Compl. ¶42 Compl. ¶187
• Accused Features: Google's use of voice-based shopping features in Gemini, where users can "go from brainstorming to browsing right within your chat" and interact with targeted ads, is accused of infringement Compl. ¶192

U.S. Patent No. 8,886,536 - SYSTEM AND METHOD FOR DELIVERING TARGETED ADVERTISEMENTS AND TRACKING ADVERTISEMENT INTERACTIONS IN VOICE RECOGNITION CONTEXTS

• Issued: November 11, 2014 (the "'536 Patent")
• Technology Synopsis: This patent claims a method for providing promotional content in a multi-device context. After an initial user utterance and system response, a second utterance can contain requests for two different devices, and the system determines and presents promotional content related to either request Compl. ¶45
• Asserted Claims: Claim 44 is asserted Compl. ¶45 Compl. ¶202
• Accused Features: Google Assistant features that trigger promotional content from third-party services, such as booking a ride-share, are accused of infringement Compl. ¶207

U.S. Patent No. 9,269,097 - SYSTEM AND METHOD FOR DELIVERING TARGETED ADVERTISEMENTS AND/OR PROVIDING NATURAL LANGUAGE PROCESSING BASED ON ADVERTISEMENTS

• Issued: February 23, 2016 (the "'097 Patent")
• Technology Synopsis: The patent describes a method where, after an advertisement is presented, the system interprets a subsequent user utterance based on that advertisement. A key feature is determining whether a pronoun in the user's speech refers to the product or service in the ad Compl. ¶48
• Asserted Claims: Claim 1 is asserted Compl. ¶48 Compl. ¶217
• Accused Features: Google's "Shop with AI Mode," which allows users to use natural language to describe products and receive targeted ads in response, is accused of infringement Compl. ¶222

U.S. Patent No. 12,236,456 - SYSTEM AND METHOD FOR DELIVERING TARGETED ADVERTISEMENTS AND/OR PROVIDING NATURAL LANGUAGE PROCESSING BASED ON ADVERTISEMENTS

• Issued: February 25, 2025 (the "'456 Patent")
• Technology Synopsis: This patent claims a method for processing voice-based utterances to select and present purchase opportunities. The system determines a context, selects a purchase opportunity, tracks user interaction with it over time, builds a user-specific profile based on that interaction, and uses the profile to interpret subsequent requests Compl. ¶51
• Asserted Claims: Claim 1 is asserted Compl. ¶51 Compl. ¶232
• Accused Features: Google's AI-powered shopping features in Gemini, which allow users to browse and purchase products using natural language, are accused of infringement Compl. ¶236 Compl. ¶237

U.S. Patent No. 11,087,385 - VOICE COMMERCE

• Issued: August 10, 2021 (the "'385 Patent")
• Technology Synopsis: The patent describes a "voice commerce" method where a product or service is selected for purchase based on a single first user input, without further input from the user. The system then completes the purchase transaction after receiving a second user input confirming the selection Compl. ¶54
• Asserted Claims: Claim 1 is asserted Compl. ¶54 Compl. ¶247
• Accused Features: Google features that allow users to complete purchases by voice on Google Express or within Gemini are accused of infringement Compl. ¶252 Compl. ¶253

III. The Accused Instrumentality

Product Identification

The complaint collectively refers to the accused instrumentalities as the "Google AI Products" Compl. ¶4 This includes, but is not limited to, Google Assistant, Gemini, Google voice search, YouTube voice search, the Android Auto platform, and the underlying software, AI models, and hardware infrastructure that power them Compl. ¶¶3-4

Functionality and Market Context

• The accused functionality is the "voice-based, natural language understanding and AI functionality" within the Google AI Products Compl. ¶5 This functionality is implemented via what the complaint terms "Specialized AI Software," which includes components like speech recognition engines, conversational language processors, and NLU models Compl. ¶5
• The complaint alleges that this software is activated when a user invokes a voice command, for example by pressing a microphone button or using a "Hey Google" wake command Compl. ¶5 Plaintiff contends this Specialized AI Software is a "separate and distinct" feature that constitutes a material part of the patented inventions and has no substantial non-infringing uses when invoked Compl. ¶5 A screenshot in the complaint shows Google's instructions for using the "Hey Google" wake word to "Continue the conversation" Compl. p. 36
• The complaint alleges these technologies are "fundamental to the development of voice-based AI systems" Compl. ¶3 and that Google's use of them "implicate major revenue streams" Compl. ¶71

IV. Analysis of Infringement Allegations

'681 Patent Infringement Allegations

Claim Element (from Independent Claim 1)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
A method for providing a cooperative conversational voice user interface, comprising:	Google AI Products, such as Google Assistant, are alleged to perform a method for providing a cooperative conversational voice user interface.	¶95; Ex. L-1	col. 1:18-21
receiving an utterance at a voice input device during a current conversation with a user, wherein the utterance includes one or more words that have different meanings in different contexts;	Google AI Products allegedly receive user utterances via audio input devices. These utterances may include ambiguous words, such as "the one for tomorrow at 6:00," where "one" refers to a previously discussed alarm.	¶95; Ex. L-1	col. 5:49-55
accumulating short-term shared knowledge about the current conversation...	Google Assistant allegedly accumulates short-term knowledge, such as using previous interactions and the currently displayed screen content to respond to follow-up questions. A screenshot of a Google support page titled "Continue the conversation" is provided as evidence.	¶100; p. 36	col. 5:10-14
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;	Google Assistant allegedly uses machine learning to train models on past user interactions and enables features like "Voice Match" and "Personal References" to provide personalized responses based on long-term user data, such as calendar events or contacts.	¶102; p. 38	col. 5:14-18
determining an intended meaning for the utterance, wherein determining the intended meaning for the utterance includes: identifying, at a conversational speech engine, a context associated with the utterance from the short-term shared knowledge and the long-term shared knowledge;	Google's conversational speech engine allegedly identifies context using both short-term conversational data and long-term user data. For example, it is alleged to use the time of day as a contextual cue for breakfast-related queries.	¶95; Ex. L-1	col. 11:21-25
and establishing the intended meaning within the identified context to disambiguate an intent that the user had in speaking the one or more words that have the different meanings in the different contexts;	The system allegedly disambiguates user intent based on context. For example, if a user asks about Miami and then says "show me the nicest beaches," the system understands this means beaches in Miami.	¶95; Ex. L-1	col. 12:1-5
and generating a response to the utterance, wherein the conversational speech engine grammatically or syntactically adapts the response based on the intended meaning established within the identified context.	Google AI Products allegedly generate adapted responses, such as confirming that an alarm has been moved in response to a contextual user command.	¶95; Ex. L-1	col. 13:4-9

• Identified Points of Contention:
  • Scope Questions: A central point of contention may be whether Google's use of conversational history and user data constitutes "accumulating short-term and long-term shared knowledge" as those terms are defined and used in the '681 Patent. Google may argue its architecture and data processing methods are fundamentally different from the patented solution.
  • Technical Questions: The infringement analysis may turn on how, precisely, Google's system "disambiguates" user intent. Plaintiff alleges this is done by identifying context from shared knowledge, while Google may argue its NLU models use a distinct, non-infringing process to determine intent.

'765 Patent Infringement Allegations

Claim Element (from Independent Claim 1)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
A method of providing a voice interface, comprising:	Google AI Products, such as Google Assistant and Google Voice Search, are alleged to perform a method for providing a voice interface.	¶111; Ex. M-1	col. 1:22-23
receiving, at a speech engine, a natural language utterance from a voice-enabled device, the natural language utterance corresponding to a conversation type;	Google AI Products allegedly receive natural language utterances from users via voice-enabled devices, such as a "good morning" query that corresponds to a didactic-type conversation.	¶116; p. 43	col. 4:18-22
determining the conversation type corresponding to the natural language utterance based on whether a user that spoke the natural language utterance has a leader role in an interaction with the voice-enabled device or has a supporter role in the interaction with the voice-enabled device;	The complaint alleges Google's system determines the conversation type based on the user's role. It presents an example where a user initiating a request to set an alarm has a "leader role," versus a user saying "good morning" and receiving a series of informational updates has a "supporter role."	¶111; Ex. M-1	col. 15:35-42
and generating a response to the natural language utterance with a format based on the conversation type, wherein the format is adapted to limit the user's future input to interjecting queries or requests for clarification if the user has the supporter role.	The complaint alleges that when the user has a supporter role (e.g., after saying "good morning"), the system provides a continuous response without prompting for more input, thereby limiting the user's input to interjections. This is contrasted with a leader-role interaction where the system may prompt for more information.	¶111; Ex. M-1	col. 16:32-38

• Identified Points of Contention:
  • Scope Questions: The case may hinge on the construction of "leader role" and "supporter role." The court will need to determine if Google's system makes a functional distinction that maps onto these claim terms, or if Google's conversational logic operates on different, non-infringing principles.
  • Technical Questions: A key evidentiary question will be whether Google's system actually "generat[es] a response with a format... adapted to limit the user's future input" when a user is in a "supporter role." Plaintiff will need to show not just a difference in conversational flow, but that this difference results from a specific, claim-required formatting of the response based on a determination of the user's role.

V. Key Claim Terms for Construction

• The Term: "shared knowledge" ('681 Patent, claim 1)

• Context and Importance: This term is the foundation of the '681 Patent's invention. The outcome of the infringement analysis will depend heavily on whether Google's methods of storing and using conversational history and user data fall within the scope of "accumulating short-term and long-term shared knowledge." Practitioners may focus on this term because it distinguishes the claimed invention from simple, stateless command systems.

• Intrinsic Evidence for Interpretation:

  • Evidence for a Broader Interpretation: The specification of the related '249 Patent states that shared knowledge can include "cross-modality awareness," such as a "current user interface state," and that long-term knowledge can include "explicit and/or implicit user preferences" and "most often used word choices" '249 Patent, col. 5:19-24 '249 Patent, col. 5:35-41 This language may support a broad definition covering a wide range of contextual data.
  • Evidence for a Narrower Interpretation: The figures in the related patents depict "Shared Knowledge" as a distinct architectural block within a "Cooperative Conversations" system '249 Patent, Fig. 3 A defendant may argue this implies a specific, structured data store, rather than any generalized use of user history or session data.

• The Term: "leader role" / "supporter role" ('765 Patent, claim 1)

• Context and Importance: These terms define the core logic of claim 1 of the '765 Patent. The infringement question will turn on whether Google's systems are shown to actually make a determination of a user's "role" and format the response accordingly. Practitioners may focus on this term because it requires proving not just that a conversation is happening, but that the system is classifying the user's function within that conversation.

• Intrinsic Evidence for Interpretation:

  • Evidence for a Broader Interpretation: The specification of the related '249 Patent describes a "query conversation" where a "leader" has a "specific goal in mind," which could be argued to cover any user-initiated request '249 Patent, col. 15:35-41
  • Evidence for a Narrower Interpretation: The patent itself, as recited in the complaint, states that the format for a supporter role is adapted to "limit the user's future input to interjecting queries or requests for clarification" Compl. ¶27 A defendant may argue this requires a very specific system behavior that is not present in its products, and that the terms cannot simply mean "the one asking" versus "the one answering."

VI. Other Allegations

• Indirect Infringement: The complaint makes detailed allegations of both induced and contributory infringement for all asserted patents.
  • Inducement: Plaintiff alleges Google actively induces infringement by providing extensive documentation, user manuals, and marketing materials that instruct and encourage users to operate Google AI Products in an infringing manner (e.g., using "Continued Conversation") (Compl. ¶¶99-102; Compl. ¶¶115-118). A screenshot from a Google support page instructing users how to "Continue the conversation" is cited as evidence Compl. p. 36
  • Contributory Infringement: Plaintiff alleges that the "Specialized AI Software" within Google's products is a material part of the invention, is especially made or adapted for practicing the patents, and is not a staple article of commerce with substantial non-infringing uses (Compl. ¶¶103-105).
• Willful Infringement: The complaint alleges willful infringement based on both pre-suit and post-suit knowledge. It alleges Google had pre-suit knowledge of the patents and their relevance as early as 2007-2012 through direct meetings, presentations, and a specific request by a Google patent attorney for VoiceBox's patent list (Compl. ¶¶56-65; Compl. ¶97). The complaint further alleges that Google has repeatedly cited the asserted patents as prior art during the prosecution of its own patent applications, demonstrating ongoing awareness (Compl. ¶¶76-87).

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

• A core issue will be one of definitional scope: for the '681 Patent, can Plaintiff prove that Google's method of handling user data and session history constitutes "accumulating short-term and long-term shared knowledge" as claimed, or will Google demonstrate that its architecture is fundamentally different?
• A central evidentiary question for the '765 Patent will be one of functional operation: does the accused Google Assistant software actually make a binary determination of a user's "leader role" versus "supporter role" to format its response, as required by the claim, or does it utilize a more complex, multi-factor, or different logic for managing conversational flow?
• A key aspect of the case, particularly concerning damages, will be knowledge and intent: given the extensive history of alleged disclosures and Google's own citations to the patents-in-suit as prior art, the court will have to resolve whether Google's development and offering of its AI Products was done with knowledge of, and willful blindness or deliberate disregard for, Plaintiff's patent rights.