Gatekeeper Solutions Inc v. Darktrace Inc

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: Gatekeeper Solutions Inc. (New York)
  • Defendant: Darktrace, Inc. (Delaware)
  • Plaintiff’s Counsel: DNL Zito
• Case Identification: 4:24-cv-00723, E.D. Tex., 08/13/2024
• Venue Allegations: Venue is alleged to be proper in the Eastern District of Texas because Defendant Darktrace maintains a place of business in Plano, Texas, within the district, and has allegedly committed acts of infringement there.
• Core Dispute: Plaintiff alleges that Defendant’s Darktrace/Email product infringes a patent related to systems and methods for preventing electronic communications from being sent to "conflicting recipients."
• Technical Context: The technology addresses the risk of misdirecting emails and other electronic communications, a common problem in corporate environments where sending sensitive information to an incorrect or adverse party can have significant consequences.
• Key Procedural History: The complaint alleges that the Defendant was made aware of the patent-in-suit prior to the lawsuit but continued its allegedly infringing activities, which forms the basis for the willfulness allegation.

Case Timeline

Date	Event
2013-09-13	U.S. Patent 9,032,038 Priority Date
2015-05-12	U.S. Patent 9,032,038 Issued
2024-08-13	Complaint filed in the Eastern District of Texas

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

• Patent Identification: U.S. Patent No. 9,032,038, "Recipient control system for ensuring non-conflicting and comprehensive distribution of digital information and method thereof," issued May 12, 2015.

The Invention Explained

• Problem Addressed: The patent's background section describes the increasing risk of misdirecting electronic communications like emails due to features like address auto-completion or simple user error. This can lead to dire consequences when sensitive or privileged information is sent to a "conflicting" party, such as a competitor, opposing counsel in a lawsuit, or an insurer when the insured is also a recipient (’038 Patent, col. 1:15-54).
• The Patented Solution: The invention is a system that intercepts an outgoing electronic communication and inspects the list of recipients against a set of pre-defined rules or parameters that identify "conflicting" relationships. If a conflict is detected (e.g., an email is addressed to both an internal manager and a known competitor), the system stops the communication from being sent and notifies the user, who may have an option to override the block (’038 Patent, Abstract; Fig. 1). The system can define conflicts based on email domains, specific email addresses, or other contact information like titles (’038 Patent, col. 5:7-21, col. 5:61-65).
• Technical Importance: The technology provides a rules-based, automated safeguard integrated into the communication workflow to prevent potentially damaging data breaches before they occur, moving beyond simple user warnings or post-delivery deletion requests (’038 Patent, col. 1:55-62).

Key Claims at a Glance

• The complaint asserts claims 1-45, with a focus on independent system claim 1 and independent method claim 23 (Compl. ¶11, ¶14).
• Independent Claim 1, a system claim, recites the following essential elements written in means-plus-function format:
  • means for receiving one or more parameters identifying conditions for a conflicting recipient for each recipient;
  • means for storing said parameters;
  • means for comparing the parameters of each recipient of said electronic communication with said parameters of other recipients to determine whether any of the other recipients is a conflicting recipient;
  • means for stopping the sending of the electronic communication when said comparing means determines at least one conflicting recipient;
  • means for notifying the user of each conflicting recipient...and the parameters that identify each conflicting recipient; and
  • means for sending the electronic communication when said comparing means does not determine at least one conflicting recipient.
• The complaint also alleges infringement of dependent claims but does not specify which ones (Compl. ¶11).

III. The Accused Instrumentality

Product Identification

• Product Identification: The accused product is the Darktrace/Email product (Compl. ¶4).

Functionality and Market Context

• The complaint alleges that Darktrace/Email is an email security product that uses artificial intelligence (AI) to provide "behavioral detections of misdirected emails" (Compl. ¶22, p. 6).
• The accused functionality involves the AI analyzing the recipients of an outgoing email, considering the "contextual relationship between the sender and recipients" and "the relationships the recipients have with each other" (Compl. ¶22, p. 6). This screenshot describes how the AI analyzes recipients to identify potential errors. The system is alleged to be "always on and continuously learning" to establish a baseline of normal communication patterns, which it then uses to flag deviations (Compl. ¶22, p. 6).
• When an anomaly is detected, the product can allegedly "block the email" or "invite the user to think twice," thereby preventing data leakage (Compl. ¶22, p. 7).

IV. Analysis of Infringement Allegations

’038 Patent Infringement Allegations

Claim Element (from Independent Claim 1)	Alleged Infringing Functionality -	Complaint Citation	Patent Citation
means for receiving one or more parameters identifying conditions for a conflicting recipient for each recipient	The AI is alleged to analyze recipients, considering "the contextual relationship between the sender and recipients, the relationships the recipients have with each other, how similar each recipient’s name and history is to other known contacts, and the names of attached files." This screenshot shows the AI analyzing contextual relationships among recipients. (Compl. ¶22, p. 6)	¶22	col. 7:48-54
means for storing said parameters	The system is described as "always on and continuously learning," allowing it to "learn the new normal, including common behaviors, recipients, attached file names, and more," which implies storing these learned parameters. (Compl. ¶22, p. 6) -	¶22	col. 7:55-61
means for comparing the parameters of each recipient...to determine whether any of the other recipients is a conflicting recipient	The accused product allegedly has a "classifier that tracks semantics shared over a network of correspondents" and "detects anomalous messages by comparing message content to that of previous communications." This screenshot alleges the product compares message content to detect anomalies. (Compl. ¶22, p. 7) -	¶22	col. 10:40-47
means for stopping the sending of the electronic communication when said comparing means determines at least one conflicting recipient	The complaint alleges that the AI "can block the email" if it determines the email is "outside of a user’s typical behavior." This screenshot describes the AI's ability to block an email. (Compl. ¶22, p. 7) -	¶22	col. 11:18-20
means for notifying the user of each conflicting recipient...and the parameters that identify each conflicting recipient	The complaint states the product "may alert the user" and can "invite the user to think twice." This screenshot indicates the AI provides alerts to the user. (Compl. ¶22, p. 7) -	¶22	col. 10:50-54
means for sending the electronic communication when said comparing means does not determine at least one conflicting recipient	The product is alleged to work "in the background to reduce risk while allowing legitimate emails to flow." This screenshot indicates that non-conflicting emails are sent normally. (Compl. ¶22, p. 8) -	¶22	col. 10:45-47

• Identified Points of Contention:
  • Scope Questions (Means-Plus-Function): A central issue will be the interpretation of the "means for..." limitations under 35 U.S.C. § 112(f). The scope of each "means" is not the functional language itself, but the specific structures disclosed in the patent's specification that perform the function, and their equivalents. The complaint alleges infringement by an "AI" that performs functions like "behavioral detections" and "unsupervised learning" (Compl. ¶22, pp. 6-7). A key question is whether the patent's specification discloses a corresponding structure (e.g., specific algorithms, database structures, software modules) that is structurally equivalent to the accused AI-based system.
  • Technical Questions: The patent describes a system based on pre-defined, user-configurable rules (e.g., creating a "conflict list" or "block list") (’038 Patent, Fig. 2, Fig. 12). The complaint, however, describes the accused product as using adaptive, AI-driven "unsupervised learning" to "learn the new normal" and detect "anomalous messages" (Compl. ¶22, pp. 6-7). This raises the question of whether an adaptive, behavior-based AI system operates in the same way as, or is an equivalent to, the more deterministic, rules-based system described and depicted in the patent's specification.

V. Key Claim Terms for Construction

• The Term: "conflicting recipient"

• Context and Importance: This term is at the heart of the invention and appears in the preamble of the asserted independent claim. Its definition dictates the scope of what the system is designed to prevent. The dispute will likely center on whether the accused product's detection of "anomalous messages" or "misdirected emails" based on learned behavioral patterns falls within the meaning of identifying a "conflicting recipient" as contemplated by the patent.

• Intrinsic Evidence for Interpretation:

  • Evidence for a Broader Interpretation: The specification provides a wide array of examples, including "an insurer and the insured; competing businesses; management and staff; parties involved in medical care who have differing interests; and opposing parties in a lawsuit" (’038 Patent, col. 1:45-50). This list suggests "conflicting" is a broad concept covering any relationship where co-receipt of information is undesirable.
  • Evidence for a Narrower Interpretation: The patent’s implementation examples and figures heavily emphasize user-defined, explicit rules. For instance, Figure 2 shows a "Conflict list (rule)" being manually created (’038 Patent, Fig. 2), and Figure 12 shows a user interface for creating rules like "Domain Block" and "Recipient Conflict" (’038 Patent, Fig. 12). This could support a narrower construction requiring a pre-defined, explicit conflict rule, rather than an inferred or behavior-based anomaly.

• The Term: "means for comparing"

• Context and Importance: This is a means-plus-function term. Its construction will determine what specific algorithm or process disclosed in the patent is covered. Practitioners may focus on this term because the complaint alleges the accused product "compares message content" using an "AI classifier" (Compl. ¶22, p. 7), while the patent may disclose a different structure.

• Intrinsic Evidence for Interpretation:

  • Evidence for a Broader Interpretation: The function itself—"comparing the parameters of each recipient"—is stated broadly. An argument could be made that any underlying software module that performs this check is a corresponding structure.
  • Evidence for a Narrower Interpretation: The corresponding structure appears to be the inspection of a distribution list against a database of pre-set rules (see Fig. 1, step 102 "Inspect distribution list" based on "Database parameters" 104; ’038 Patent, col. 10:40-47). The court may be asked to decide if this disclosed structure is limited to checking recipient lists against pre-defined conflict pairs, as opposed to the accused product's alleged function of "comparing message content" using "semantics" (’038 Patent, Fig. 10; Compl. ¶22, p. 7).

VI. Other Allegations

• Indirect Infringement: The complaint alleges active inducement by asserting that Darktrace provides software and instructions to its customers, enabling and encouraging them to use the Darktrace/Email product in an infringing manner (Compl. ¶17, ¶26). Contributory infringement is alleged on the basis that Darktrace provides a material part of the infringing system that is not a staple article of commerce and is known to be especially adapted for infringement (Compl. ¶18, ¶27).
• Willful Infringement: Willfulness is alleged based on the claim that Defendant was made aware of the ’038 Patent and its infringement but "refused to cease its infringing actions" and continued to infringe intentionally (Compl. ¶1, ¶29).

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

This case will likely focus on fundamental questions of claim scope and technical operation, particularly given the patent’s use of means-plus-function claiming. The central issues for the court will be:

1. A core issue will be one of structural correspondence under § 112(f): Does the patent’s specification disclose an algorithm or software structure that is the same as or equivalent to the accused "AI," "unsupervised learning," and "classifier" technology that allegedly performs the claimed functions of comparing, stopping, and notifying?

2. A second key issue will be one of definitional scope: Can the term "conflicting recipient", which the patent illustrates with examples of explicit, pre-defined adverse relationships, be construed to cover the "anomalous messages" and "behavioral detections of misdirected emails" identified by the accused product’s adaptive AI?

3. A final evidentiary question will be one of functional operation: Does the accused Darktrace/Email system, which allegedly learns from user behavior to detect anomalies, operate in a fundamentally different way from the patent’s disclosed system, which relies on inspecting recipient lists against user-configured databases of static conflict rules?