DCT
2:25-cv-15207
Kabir v. Webmd LLC
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: Azad Alamgir Kabir, M.D., MSPH (pro se)
- Defendant: WebMD LLC, THOUGHTi Inc., and JOHN DOE 1 (believed to be OpenAI, Inc.)
- Plaintiff’s Counsel: Pro Se
- Case Identification: 2:25-cv-15207, D.N.J., 10/06/2025
- Venue Allegations: Plaintiff alleges venue is proper because all Defendants conduct substantial business in the District of New Jersey.
- Core Dispute: Plaintiff alleges that Defendants' WebMD Symptom Checker tool infringes two patents related to automated medical diagnosis systems, and further alleges this infringement stems from the misappropriation of Plaintiff's proprietary database and algorithms by a former contractor.
- Technical Context: The technology relates to computer-implemented diagnostic tools that process patient-reported symptoms to generate a ranked list of potential medical conditions, a key feature in modern consumer-facing digital health platforms.
- Key Procedural History: The filing is a Third Amended Complaint. The patents-in-suit share a common specification, with the later-issued patent being a continuation of the earlier one. The complaint alleges a specific history of misconduct, asserting that Defendant THOUGHTi, Inc. gained access to Plaintiff’s proprietary system in 2017 under a non-disclosure agreement and subsequently transferred the technology to Defendant WebMD.
Case Timeline
| Date | Event |
|---|---|
| 2012-07-25 | Earliest Priority Date for ’865 and ’051 Patents |
| 2017-01-03 | U.S. Patent No. 9,536,051 Issues |
| 2017 | Plaintiff engages THOUGHTi, Inc. |
| 2017–2018 | THOUGHTi allegedly transfers Plaintiff's technology to WebMD |
| 2020 | Plaintiff discovers WebMD's Symptom Checker |
| 2024-04-30 | U.S. Patent No. 11,972,865 Issues |
| 2025-10-06 | Third Amended Complaint Filed |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 11,972,865 B1 - "High Probability Differential Diagnoses Generator And Smart Electronic Medical Record"
The Invention Explained
- Problem Addressed: The patent’s background section describes the high risk of diagnostic error in medicine, particularly when inexperienced clinicians formulate initial hypotheses early in a patient encounter without a structured, efficient methodology. Existing electronic medical records are described as inefficient and lacking integrated tools to aid in diagnostic decision-making (’865 Patent, col. 1:49-66).
- The Patented Solution: The invention is a computer-implemented method that receives multiple patient symptoms, compares this input against a medical database linking symptoms to diseases, and generates a short, ranked list of high-probability differential diagnoses. This automated process is designed to help a healthcare provider quickly focus on the most likely conditions (’865 Patent, Abstract; col. 6:26-52).
- Technical Importance: The system aims to improve diagnostic accuracy and healthcare efficiency by providing a rapid, data-driven framework for generating initial diagnostic possibilities (’865 Patent, col. 3:41-52).
Key Claims at a Glance
- The complaint does not identify specific claims asserted against the Defendants. Independent claim 1 is analyzed here as a representative claim.
- Essential Elements of Independent Claim 1:
- Collecting a first and second medical clinical data from a patient.
- Linking the collected data with a differential diagnosis medical database.
- Comparing the collected data to the database to isolate common diseases and arrange them into distinct groupings for each data point.
- Isolating disease data common across the different groupings.
- Generating a list of the common disease data.
- Ranking the list based on (1) the number of times a disease is associated with the input data and (2) the disease's pre-existing relative rank for that data.
- The complaint does not explicitly reserve the right to assert dependent claims.
U.S. Patent No. 9,536,051 B1 - "High Probability Differential Diagnoses Generator"
The Invention Explained
- Problem Addressed: As a parent to the ’865 Patent with a shared specification, the ’051 Patent addresses the same problem: the need for a tool to help healthcare providers process multiple symptoms to generate a clinically meaningful list of potential diagnoses, thereby reducing error and inefficiency (’051 Patent, col. 2:16-22).
- The Patented Solution: The invention provides an automated method for generating a differential diagnosis by comparing patient data against a medical database. The system is designed to produce a ranked list of high-probability diagnoses to assist providers in treating patients (’051 Patent, Abstract; col. 4:45-51).
- Technical Importance: The technology provides a systematic approach to diagnostic reasoning, which is particularly valuable for training environments and for reducing reliance on memory or unstructured hypothesis generation (’051 Patent, col. 2:65-col. 3:7).
Key Claims at a Glance
- The complaint does not identify specific claims asserted against the Defendants. Independent claim 1 is analyzed here as a representative claim.
- Essential Elements of Independent Claim 1:
- Providing a selection means for a user to select first and second medical clinical data.
- Comparing the selected data to a medical database.
- Grouping disease data from the database in response to the comparison.
- Ranking the grouped disease data.
- Illustrating a listing of the ranked grouped disease data.
- The complaint does not explicitly reserve the right to assert dependent claims.
III. The Accused Instrumentality
Product Identification
- The accused instrumentality is Defendant WebMD, LLC's "Symptom Checker" tool (Compl. ¶7, ¶14).
Functionality and Market Context
- The complaint alleges the Symptom Checker tool "integrates backend logic" from Defendant OpenAI, Inc., which powers a model identified as "saas-openai-gpt-4o-mini" (Compl. ¶18A). This logic is accused of "generating ranked differential diagnoses using symptom-to-condition scoring," which produces outputs showing "strength-of-match values, tiered match labels (e.g., 'High Match'), and dynamically generated lists of candidate conditions" (Compl. ¶18B). The complaint includes Exhibit 44, described as a "WebMD API JSON Payload Captured via Browser Inspection," which allegedly provides evidence of this functionality (Compl. ¶20; Ex. 44). The tool is alleged to be a significant revenue generator for WebMD (Compl. ¶17).
IV. Analysis of Infringement Allegations
The complaint does not contain a claim chart or specify which claims are asserted. The following charts are constructed based on the general infringement theory presented in the complaint as applied to representative independent claim 1 of each patent.
U.S. Patent No. 11,972,865 Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| collecting a first medical clinical data from a patient; collecting a second medical clinical data from the patient... | The WebMD Symptom Checker tool prompts users to input their symptoms to initiate a diagnosis. | ¶14, ¶18B | col. 6:50-52 |
| linking said collected first and second medical clinical data with a differential diagnosis medical database... | The tool is alleged to use a backend system, including an AI model and potentially supplemental logic or databases, to process user-inputted symptoms. | ¶2, ¶18A | col. 5:42-52 |
| isolating all disease data common to said first medical clinical data... isolating all disease data common to said second medical clinical data... | The complaint alleges the system performs "symptom-to-condition scoring" to map input symptoms to potential diseases. | ¶18B | col. 9:10-15 |
| ranking said isolated disease data common to said first grouping and said second grouping into a third ranked list... | The accused tool is alleged to generate "ranked differential diagnoses," "strength-of-match values," and "tiered match labels" as output. | ¶2, ¶18B | col. 10:11-20 |
U.S. Patent No. 9,536,051 Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| providing a selection means whereby a first medical clinical data of a patient may be selected... | The WebMD Symptom Checker provides an interface for users to select or input their medical symptoms. | ¶14, ¶18B | col. 7:40-49 |
| comparing said first medical clinical data and said second medical clinical data to a medical database... | The system allegedly processes the input symptoms against backend logic and data sources, including an AI model. | ¶2, ¶18A | col. 8:1-3 |
| ranking said grouped disease data... | The accused tool generates "ranked differential diagnoses" and "strength-of-match values." | ¶2, ¶18B | col. 9:44-50 |
| illustrating a listing of said ranked grouped disease data... | The Symptom Checker presents users with a "dynamically generated list of candidate conditions." | ¶18B | col. 9:51-53 |
- Identified Points of Contention:
- Scope Questions: A primary issue may be whether the accused system, which allegedly utilizes a general-purpose large language model ("saas-openai-gpt-4o-mini"), meets the claim limitation of a "differential diagnosis medical database" (’865 Patent, Claim 1). The complaint itself suggests that such models are not "inherently capable" of this function without "supplemental logic modules or structured databases" (Compl. ¶2), raising the question of whether the accused product's architecture matches that described in the patents.
- Technical Questions: The infringement analysis will likely focus on whether the alleged "symptom-to-condition scoring" (Compl. ¶18B) of the accused tool performs the specific, multi-step ranking algorithm recited in claim 1 of the ’865 patent, which is based on the "number of times" a disease is associated with inputs and its "relative ranked position." The complaint does not provide sufficient detail to analyze the precise mechanism of the accused system's ranking logic.
V. Key Claim Terms for Construction
The Term: "differential diagnosis medical database" (from ’865 Patent, Claim 1).
Context and Importance: The definition of this term is central to whether the accused AI-powered system infringes. The dispute may turn on whether a system invoking a large language model can be considered a "database" as contemplated by the patent, or if that term requires a more traditionally structured, curated set of data linking symptoms to diseases.
Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The specification mentions accessing "external databases" and querying from "any data source," which could support an argument that the term is not limited to a single, static internal database (’865 Patent, col. 6:7-22).
- Evidence for a Narrower Interpretation: The patent’s figures and detailed description illustrate a system based on discrete, pre-defined relationships between specific "signs/symptoms/findings (SSF)" and "differential diagnoses (DD)," which suggests a structured, relational database rather than the probabilistic output of a generative AI model (’865 Patent, Fig. 2; col. 9:10-20).
The Term: "ranking... based upon... the number of times said disease data is associated with said first medical data and said second medical data followed by the relative ranked position" (from ’865 Patent, Claim 1).
Context and Importance: This term defines the core patented algorithm. Infringement will depend on whether the accused "symptom-to-condition scoring" (Compl. ¶18B) operates according to this specific two-factor logic.
Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: A party could argue that any system that considers both the frequency of association and the importance of a symptom-disease link performs a "ranking" that falls within the general scope of the claim.
- Evidence for a Narrower Interpretation: The claim language recites a specific sequence of operations: a primary ranking criterion (the count of matching symptoms) followed by a secondary, tie-breaking criterion ("relative ranked position"). This suggests a precise, deterministic algorithm that may be technically distinct from the statistical weighting or probability calculations of an AI model.
VI. Other Allegations
- Indirect Infringement: The complaint accuses "JOHN DOE 1 (Believed to be OpenAI, Inc.)" of indirect infringement, alleging it is a "co-conspirator" and "joint tortfeasor" for contributing to or inducing WebMD's infringement by providing the backend AI model (Compl. ¶18C). The claim is based on the allegation that OpenAI knew or was willfully blind to its model being used in an infringing manner (Compl. ¶21.I).
- Willful Infringement: Plaintiff alleges that Defendants' infringement is "willful and intentional" (Compl. ¶21.III). While no pre-suit notice is mentioned, the complaint's narrative of trade secret misappropriation via Defendant THOUGHTi could be used to argue that WebMD had knowledge of the patented technology (Compl. ¶10-13).
VII. Analyst’s Conclusion: Key Questions for the Case
- Definitional Scope: A core issue will be whether the patent claims, which appear to describe a system using a structured "database" and a specific rule-based ranking algorithm, can be construed to cover the accused functionality allegedly provided by a general-purpose large language model. The case may test the boundaries of how patent claims drafted before the rise of generative AI apply to such technology.
- Evidentiary Proof: A key question will be one of technical evidence: what proof can the Plaintiff obtain through discovery to demonstrate that the internal workings of the WebMD Symptom Checker's AI-powered "scoring" mechanism perform the specific steps of isolating, grouping, and multi-factor ranking as required by the asserted claims?
- Factual Causation: The allegations of trade secret misappropriation are central to the narrative. A critical question for the litigation will be whether Plaintiff can factually establish a chain of events where technology embodying the patented methods was improperly transferred from Plaintiff to THOUGHTi and then to WebMD, as this would be highly relevant to the issues of knowledge and willfulness.