DCT
1:23-cv-00378
Dialect LLC v. Google LLC
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: Dialect, LLC (Texas)
- Defendant: Google LLC (Delaware)
- Plaintiff’s Counsel: Farnan LLP; BLUE PEAK LAW GROUP Group
- Case Identification: 1:23-cv-00378, D. Del., 06/23/2023
- Venue Allegations: Venue is alleged to be proper in the District of Delaware because Defendant Google LLC is a Delaware limited liability company and is therefore domiciled in the district.
- Core Dispute: Plaintiff alleges that Defendant’s Google Assistant platform and associated products and services infringe eight patents related to natural language understanding and voice recognition technology.
- Technical Context: The technology concerns systems for processing natural language speech, enabling more conversational and context-aware interactions between humans and smart devices like phones, speakers, and vehicle infotainment systems.
- Key Procedural History: The complaint alleges a long history of pre-suit knowledge by Google, beginning with direct notice from the original inventor, VoiceBox Technologies, as early as 2012. It further alleges that Google repeatedly cited several of the asserted patents as prior art during the prosecution of its own patent applications, and that Google developed its Assistant platform after reviewing the asserted patents. These allegations form the basis for a claim of willful infringement.
Case Timeline
| Date | Event |
|---|---|
| 2002-06-03 | Earliest Priority Date for all Asserted Patents |
| 2008-07-08 | U.S. Patent No. 7,398,209 Issues |
| 2009-03-10 | U.S. Patent No. 7,502,738 Issues |
| 2009-12-29 | U.S. Patent No. 7,640,160 Issues |
| 2010-04-06 | U.S. Patent No. 7,693,720 Issues |
| 2011-09-06 | U.S. Patent No. 8,015,006 Issues |
| 2011-12-01 | Google allegedly informs VoiceBox of investment in similar technologies |
| 2012-08-22 | VoiceBox allegedly provides Google specific notice of seven of the patents/applications |
| 2013-05-20 | Patent application for what would become the ’652 Patent is filed |
| 2013-05-21 | U.S. Patent No. 8,447,607 Issues |
| 2014-09-07 | Google allegedly discloses the ’720 Patent as prior art to its own patent application |
| 2014-09-30 | U.S. Patent No. 8,849,652 Issues |
| 2015-05-12 | U.S. Patent No. 9,031,845 Issues |
| 2015-05-27 | Google allegedly discloses the ’607 Patent as prior art to its own patent application |
| 2016-02-04 | Google allegedly discloses the ’607 Patent as prior art to its own patent application |
| 2016-01-01 | Alleged launch year of the Google Assistant Platform |
| 2023-06-23 | Amended Complaint Filed |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 7,398,209 - "Systems And Methods For Responding To Natural Language Speech Utterance"
- Issued: July 8, 2008
The Invention Explained
- Problem Addressed: The patent’s background section describes the fundamental incompatibility between human and machine communication; human language relies heavily on context and domain knowledge, whereas machine-based queries are "highly structured and are not inherently natural to the human user" (’209 Patent, col. 1:27-35).
- The Patented Solution: The invention proposes a complete speech-based environment that uses context, user profiles, and "domain agents" to interpret natural language utterances (’209 Patent, Abstract). These agents are described as autonomous, executable modules that organize domain-specific information and behavior, allowing the system to parse a user's request, invoke the proper resources, and present results in a natural manner (’209 Patent, col. 2:48-59).
- Technical Importance: The use of modular "domain agents" represented an architectural approach to making voice systems more flexible and context-aware, moving beyond rigid, predefined command structures (’209 Patent, col. 2:48-59).
Key Claims at a Glance
- The complaint asserts at least independent Claim 1 (Compl. ¶59).
- Claim 1 of the ’209 Patent recites the essential elements of a method for responding to a speech utterance, including:
- receiving the user's speech utterance containing a request;
- maintaining a dynamic set of prior probabilities or fuzzy possibilities;
- recognizing words and phrases using dictionaries;
- parsing the words to determine a meaning and context for the request;
- selecting at least one domain agent—an autonomous executable—based on the determined meaning;
- formulating the request in accordance with a grammar used by the selected agent;
- invoking the selected domain agent to process the request; and
- presenting the results to the user.
- The complaint reserves the right to assert additional claims (Compl. ¶78).
U.S. Patent No. 7,502,738 - "Systems And Methods For Responding To Natural Language Speech Utterance"
- Issued: March 10, 2009
The Invention Explained
- Problem Addressed: Similar to its parent ’209 Patent, the ’738 Patent addresses the technical problem of making machine processing of queries compatible with natural human language, which relies heavily on context and domain knowledge (’738 Patent, col. 1:26-37).
- The Patented Solution: The ’738 Patent claims a system architecture that builds upon the "domain agent" concept. It explicitly claims an "agent architecture" comprising a plurality of domain agents, a parser, and an "event manager" to coordinate interactions (’738 Patent, Claim 1). Critically, it also claims an "update manager" that "enables the user to purchase one or more domain agents from a third party on a one-time or subscription basis," which is used to "add new agents to the system" (’738 Patent, Claim 1; col. 2:63-67).
- Technical Importance: This patent's concept of an extensible architecture where users can acquire new, third-party "domain agents" prefigures the "skills" and "actions" ecosystems of modern digital assistants (’738 Patent, Claim 1).
Key Claims at a Glance
- The complaint asserts at least independent Claim 1 (Compl. ¶97).
- Claim 1 of the ’738 Patent recites the essential elements of a system, including:
- an agent architecture that includes a plurality of domain agents, each being an autonomous executable;
- a parser configured to determine context and meaning to select a domain agent;
- an event manager to coordinate interaction between the parser and the agent architecture; and
- an update manager that enables the user to purchase one or more domain agents from a third party.
- The complaint reserves the right to assert additional claims (Compl. ¶119).
Multi-Patent Capsule: U.S. Patent No. 7,640,160
- Patent Identification: U.S. Patent No. 7,640,160, "Systems And Methods For Responding To Natural Language Speech Utterance," issued December 29, 2009 (Compl. ¶21).
- Technology Synopsis: The ’160 Patent describes a method for interpreting utterances using a "knowledge-enhanced speech recognition engine" designed to determine user intent and correct false recognitions (Compl. ¶23). The method involves identifying matching contexts by comparing text combinations against grammar entries, scoring the contexts, and selecting the highest-scoring context to determine user intent (Compl. ¶24).
- Asserted Claims: At least independent Claim 12 is asserted (Compl. ¶128).
- Accused Features: Google Assistant's alleged ability to determine the best interpretation of ambiguous queries by parsing user intent, identifying semantics, and ranking possible answers is accused of infringement (Compl. ¶130, ¶134, ¶136, ¶138).
Multi-Patent Capsule: U.S. Patent No. 7,693,720
- Patent Identification: U.S. Patent No. 7,693,720, "Mobile Systems And Methods For Responding To Natural Language Speech Utterance," issued April 6, 2010 (Compl. ¶25).
- Technology Synopsis: The ’720 Patent is directed to a mobile system, specifically for use in a vehicle, that responds to natural language speech (Compl. ¶27). The system uses a speech recognition engine with dictionaries that are "dynamically updated based on at least a history of a current dialog and one or more prior dialogs associated with the user" (Compl. ¶28).
- Asserted Claims: At least independent Claim 1 is asserted (Compl. ¶159).
- Accused Features: The complaint accuses Google Assistant for Android Auto and Android Automotive OS, which provide voice control in a vehicular context and allegedly use dialog history to interpret user intent (Compl. ¶53, ¶161, ¶167, ¶169).
Multi-Patent Capsule: U.S. Patent No. 8,015,006
- Patent Identification: U.S. Patent No. 8,015,006, "Systems And Methods For Processing Natural Language Speech Utterances With Context-Specific Domain Agents," issued September 6, 2011 (Compl. ¶29).
- Technology Synopsis: The ’006 Patent claims a method for processing speech by formulating a request according to a domain agent's grammar (Compl. ¶32). This includes determining required and optional values, extracting parameters, inferring further parameters using "a dynamic set of prior probabilities," and transforming these into formatted tokens compatible with the agent's grammar (Compl. ¶32).
- Asserted Claims: At least independent Claim 5 is asserted (Compl. ¶192).
- Accused Features: Google Assistant's alleged process of determining context, formulating requests with specific structures (e.g., for weather queries), extracting parameters, and inferring context from previous interactions is accused of infringement (Compl. ¶200, ¶202, ¶206, ¶208).
Multi-Patent Capsule: U.S. Patent No. 8,447,607
- Patent Identification: U.S. Patent No. 8,447,607, "Mobile Systems And Methods Of Supporting Natural Language Human-Machine Interactions," issued May 21, 2013 (Compl. ¶33).
- Technology Synopsis: The ’607 Patent describes a device that generates a speech transcription based on a "personal cognitive model" (from the specific user's interaction history) and a "general cognitive model" (from a plurality of users' interactions) (Compl. ¶35-36). It then identifies a matching entry in a "context stack" to identify a domain agent and determine the user's request (Compl. ¶36).
- Asserted Claims: At least independent Claim 14 is asserted (Compl. ¶233).
- Accused Features: Google Assistant's alleged use of "Voice Match" to identify users and its use of previous interactions from both the specific user and a plurality of users to generate transcriptions and determine context are accused of infringement (Compl. ¶239, ¶241).
Multi-Patent Capsule: U.S. Patent No. 8,849,652
- Patent Identification: U.S. Patent No. 8,849,652, "Mobile Systems And Methods Of Supporting Natural Language Human-Machine Interactions," issued September 30, 2014 (Compl. ¶37).
- Technology Synopsis: The ’652 Patent describes a system for processing utterances where recognized words alone are insufficient to determine the command (Compl. ¶40). The system generates a "first context stack" on a first device and synchronizes it with a "second context stack" on a second device, using the "updated context information" to determine the command (Compl. ¶40).
- Asserted Claims: At least independent Claim 1 is asserted (Compl. ¶270).
- Accused Features: The accused features include Google Assistant's ability to process ambiguous requests and its functionality for syncing activity and context across multiple devices within a Google Home environment (Compl. ¶272, ¶276, ¶282).
Multi-Patent Capsule: U.S. Patent No. 9,031,845
- Patent Identification: U.S. Patent No. 9,031,845, "Mobile Systems And Methods For Responding To Natural Language Speech Utterance," issued May 12, 2015 (Compl. ¶41).
- Technology Synopsis: The ’845 Patent claims a mobile system located at a vehicle that processes a natural language utterance and makes a determination whether to execute the resulting command or query "on-board" the vehicle or "off-board the vehicle" by invoking a device that communicates wirelessly over a wide area network (Compl. ¶44).
- Asserted Claims: At least independent Claim 1 is asserted (Compl. ¶303).
- Accused Features: Google Assistant with Android for Cars is accused of infringement, specifically its alleged ability to determine whether to execute commands on-board (e.g., media playback) or off-board by connecting to the internet (e.g., to search for charging stations) (Compl. ¶315, ¶317, ¶319).
III. The Accused Instrumentality
Product Identification
- The accused instrumentalities are collectively identified as the "Accused Google Assistant Products and Services" (Compl. ¶53). This includes the Google Assistant Platform, its various "Actions" (Conversational, App, smart home, media), Google Assistant for Android Auto, Android Automotive OS, and Dialogflow virtual agents, as well as the hardware (e.g., smartphones, Google Home devices) and software that run these services (Compl. ¶53).
Functionality and Market Context
- The complaint describes the Google Assistant as a digital platform for providing internet services, advertising, and internet-enabled hardware and software (Compl. ¶52-53). Functionally, it is alleged to be a voice-controlled system that receives natural language speech from a user, processes the speech to determine meaning and context, selects an appropriate application or service (a first-party function like Search or a third-party "Action"), formulates and processes a request, and presents a result to the user (Compl. ¶61-77). The platform is also alleged to allow third-party developers to create and offer voice-controlled applications, including through paid subscriptions via the Google Play store (Compl. ¶101, ¶107). The complaint also specifically identifies the automotive-focused versions of the platform, such as Android Auto, as accused instrumentalities (Compl. ¶53, ¶158).
IV. Analysis of Infringement Allegations
’209 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| A method responsive to a user generated natural language speech utterance, comprising: receiving the user generated natural language speech utterance... | Google Assistant is described as responding to and receiving user-generated natural language speech utterances containing requests, such as "What's the weather today?" | ¶60-63 | col. 2:20-23 |
| maintaining a dynamic set of prior probabilities or fuzzy possibilities usable at each stage of processing... | Google Assistant allegedly analyzes text in combination with "useful information such as recent requests" to interpret the utterance. | ¶64-65 | col. 4:14-23 |
| recognizing words and phrases contained in the received utterance using information in one or more dictionary and phrase tables; | The system's Natural Language Understanding (NLU) allegedly uses "trainingPhrases[]" and custom-defined phrases to automatically match words and phrases from user input. | ¶66-67 | col. 10:52-56 |
| parsing the recognized words and phrases to determine a meaning... wherein determining the meaning includes determining a context... based on one or more keywords... | The complaint alleges Google Assistant's NLU parses text to identify "the semantics, i.e. the meaning, of your question" and determine context based on keywords to select an appropriate intent. A diagram showing utterances like "What's the forecast" being matched to a "Forecast Intent" is provided as evidence (Compl. p. 26). | ¶68-69 | col. 4:46-54 |
| selecting at least one domain agent based on the determined meaning, the selected domain agent being an autonomous executable... | Google Assistant is alleged to select an application, such as Maps or Search, to respond to a request based on the determined meaning and context. | ¶70-71 | col. 6:5-7 |
| formulating the at least one request... in accordance with a grammar used by the selected domain agent... | The complaint alleges Google Assistant formulates requests using specific structures or grammars, providing the example of a weather query that extracts time and location parameters. | ¶72-73 | col. 12:54-65 |
| invoking the selected domain agent to process the formulated request; and | Google Assistant is described as invoking applications like Google Maps to process a query and return results. | ¶74-75 | col. 12:60-61 |
| presenting results of the processed request to the user... | The system is alleged to present results "in the appropriate format for your device," such as a spoken answer or a map with walking directions. | ¶76-77 | col. 4:8-10 |
’738 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| an agent architecture that includes a plurality of domain agents, each of the plurality of domain agents being an autonomous executable... | Google Assistant allegedly provides an architecture that allows third-party developers to create applications ("App Actions") that function as domain agents in a variety of contexts such as "Productivity" or "Games." A screenshot shows Google Assistant linking to a specific function within a PayPal app (Compl. p. 36). | ¶100-101 | col. 2:47-56 |
| a parser configured to determine a context... and to determine a meaning... wherein the parser selects at least one of the plurality of domain agents based on the determined meaning... | The Assistant's NLU allegedly functions as a parser that determines context and meaning from a user's utterance to select an appropriate domain agent, such as Google Maps for a location-based question. | ¶102-103 | col. 10:9-25 |
| an event manager configured to coordinate interaction between the parser and the agent architecture; and | The complaint provides a diagram described as showing how Google Assistant includes software that coordinates interactions between system components, managing the flow of a request from the user to the Assistant and then to the third-party application (Compl. p. 39). | ¶104-105 | col. 10:1-8 |
| an update manager that enables the user to purchase one or more domain agents from a third party on a one-time or subscription basis. | Google Assistant has allegedly allowed app makers to "sell subscriptions directly to users" since 2018 through the Google Play store, which is alleged to function as the claimed update manager (Compl. p. 40). | ¶106-107 | col. 2:63-67 |
Identified Points of Contention
- Scope Questions: A central question for both patents will be whether Google's "Actions" or third-party applications, which are integrated via APIs, meet the definition of a "domain agent" as an "autonomous executable" and a "complete...package of functionality" as described in the patents (’209 Patent, col. 2:50-57; ’738 Patent, Claim 1). Google may contend that its service-based architecture is fundamentally different from the modular, self-contained agent architecture disclosed.
- Technical Questions: For the ’738 Patent, a key dispute may arise over whether the Google Play store, a general-purpose application marketplace, functions as the claimed integrated "update manager" for purchasing "domain agents". Similarly, it raises the question of whether the software coordinating API calls between the Assistant and third-party apps performs the role of the claimed "event manager".
V. Key Claim Terms for Construction
The Term: "domain agent"
- Context and Importance: This term appears in the independent claims of both lead patents and is foundational to the infringement case. The viability of the plaintiff's theory depends on whether Google's "Actions" and integrated third-party applications fall within the scope of this term.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The claims define the term functionally as an "autonomous executable that receives, processes, and responds to requests associated with the determined context" (’209 Patent, Claim 1). Plaintiff may argue that any third-party service invoked by the Assistant to handle a specific domain-based request meets this functional definition.
- Evidence for a Narrower Interpretation: The specification describes agents as "complete, convenient and re-distributable packages or modules of functionality" that contain "executable code, scripts, links to information, and other forms of communication data" (’209 Patent, col. 2:50-57). This language, along with architectural diagrams showing agents as distinct blocks within the system, may support a narrower construction of a self-contained software object rather than a service invoked via an API (’738 Patent, Fig. 2).
The Term: "update manager that enables the user to purchase one or more domain agents from a third party"
- Context and Importance: This limitation from Claim 1 of the ’738 Patent is highly specific. Infringement of this claim hinges on mapping this element to a component of the accused Google ecosystem. Practitioners may focus on this term because it appears to describe a commercial "app store" model for voice capabilities.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The claim uses broad, functional language: a manager that "enables the user to purchase." Plaintiff will likely point to allegations that Google Assistant allows developers to sell subscriptions to their "Actions" through the Google Play store as direct evidence of this functionality (Compl. ¶107).
- Evidence for a Narrower Interpretation: The patent describes the "update manager" as an integrated component of the claimed system architecture, alongside the parser and event manager, which is "used to add new agents to the system" (’738 Patent, col. 2:63-67; Fig. 2). Defendant may argue that the Google Play store is a separate, pre-existing, general-purpose digital storefront, not the integrated "update manager" component claimed in the patent.
VI. Other Allegations
Indirect Infringement
- The complaint alleges active inducement of infringement under 35 U.S.C. § 271(b) (Compl. ¶82, ¶111). The basis for this allegation is that Google provides the Accused Google Assistant Products and Services and encourages and instructs businesses (developers) and consumers to use them in their "ordinary, customary, and intended way," which allegedly constitutes direct infringement (Compl. ¶84, ¶113).
Willful Infringement
- Willfulness is alleged for all asserted patents. The complaint bases this on Google’s alleged pre-suit knowledge of the patents, dating back to at least 2012 from direct communications with the original assignee, VoiceBox (Compl. ¶45-46). The complaint further supports this allegation by asserting that Google repeatedly cited the asserted patents as prior art during the prosecution of its own patent applications and developed the accused platform in light of its review of the patents (Compl. ¶49, ¶51, ¶90).
VII. Analyst’s Conclusion: Key Questions for the Case
- A core issue will be one of architectural equivalence: can the components of Google's distributed, service-based ecosystem (including the Assistant platform and the separate Google Play store) be mapped onto the more integrated, modular architecture described in the patents, which claims specific components like "domain agents", an "event manager", and an "update manager"?
- A key question of definitional scope will be whether the term "domain agent", described in the specification as a self-contained "package of executable code," can be construed broadly enough to read on the "Actions" and third-party applications that are invoked by Google Assistant through APIs.
- A central evidentiary question will concern willfulness: given the detailed allegations of direct notice and prior art citations dating back over a decade, the case will likely focus on whether Google's conduct was "deliberate or intentional," potentially exposing it to enhanced damages if infringement is found.