# US 12,602,459 B2 · Systems and methods for chatbot authentication > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,602,459 B2 - Document ID: US12602459B2 - Status: Granted - Kind: B2 - Application number: 18/913,230 - Filed: 2024-10-11 - Published: 2025-05-08 - Granted: 2026-04-14 - Earliest claimed: 2023-11-08 - Assignees: LexisNexis Risk Solutions Inc. - Applicants: LexisNexis Risk Solutions, Inc. - Inventors: Lancelot Prégniard; Sharfaraz Salek; Yaniv Koss; Peter Greene; Philip Michael Flynn - Web page: https://patentgrep.com/patent/US12602459B2 ## CPC - G06F 21/31: User authentication - G06F 21/316: {by observing the pattern of computer usage, e.g. typical user behaviour} - H04L 51/02: using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages - G06F 2221/2103: Challenge-response - G06F 2221/2135: Metering - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) ## Abstract Systems and methods for a chatbot authentication system and process that may utilize a large language model (LLM) and state of the art generative natural language processing (NLP) to generate and/or interpret user answers questions generated based on comprehensive known data about the user. The questions can include skills-based and/or knowledge-based queries which may be open ended questions. The user answers provided may be evaluated for accuracy and completeness. Iterative and/or follow-up questions may be generated and posed to allow the user to provide clarifications or additional details. By virtue of the LLM and NLP, the systems and methods disclosed herein may understand the semantical meaning of the answers provided by the user, which can enable a more accurate, user friendly and secure authentication process. ## Claims ### Claim 1 A computer-implemented method for user authentication for access to a service using a chatbot, the method comprising: receiving user information corresponding to a user; resolving the received user information to a unique identifier (UID) for the user; obtaining, from one or more data sources, comprehensive data about the user that matches the UID; generating a plurality of open-ended authentication questions based on the comprehensive data; outputting for display on a user device associated with the user, one or more of the plurality of open-ended authentication questions; receiving, in a natural language, one or more user answers corresponding to the one or more of the plurality of open-ended authentication questions; utilizing a large language model (LLM) to evaluate the one or more user answers against a known ground truth based on the comprehensive data; and responsive to determining that a success count corresponding to a number of factually correct user answers corresponding to the one or more of the plurality of open-ended authentication questions matches a pre-defined threshold number of matches, authenticating the user for access to the service. ### Claim 2 (depends on 1) The computer-implemented method of claim 1, further comprising: gathering user device data and behavioral data responsive to the user accessing an authentication page; and generating a digital profile associated with the user. ### Claim 3 (depends on 1) The computer-implemented method of claim 1, wherein generating the plurality of open-ended authentication questions is further based on a pre-defined configuration. ### Claim 4 (depends on 3) The computer-implemented method of claim 3, wherein the pre-defined configuration comprises one or more of: a number of open-ended authentication questions to generate; a threshold number of matches required for authentication, wherein the threshold number is less than or equal to the number of open-ended authentication questions; a complexity level of the open-ended authentication questions to generate, wherein in the complexity level comprises a number of contextual variables related to the user; and a level of detail required in the one or more user answers. ### Claim 5 (depends on 1) The computer-implemented method of claim 1, wherein a LLM is utilized for generating the plurality of open-ended authentication questions. ### Claim 6 (depends on 1) The computer-implemented method of claim 1, wherein: if a user answer is evaluated as factually correct, outputting for display on the user device, a next question and incrementing the success count; if a user answer is evaluated as factually incorrect, outputting for display on the user device, a next question without incrementing the success count; if a user answer is factually correct but incomplete, outputting for display on the user device, additional instructions to indicate missing information without incrementing the success count; and if a user answer is incorrect and incomplete, outputting for display on the user device, additional instructions to indicate missing information without incrementing the success count. ### Claim 7 (depends on 1) The computer-implemented method of claim 1, wherein the one or more data sources comprise one or more of: court records, birth certificate records, census bureau records, county records, property records, real estate records, court records, business records, school records, education records, webpages, professional license records, historical records, social media records, marketing records, publications, private records, media records, background records, proprietary data, and derived data records. ### Claim 8 A computer-implemented method for user authentication for access to a service using a chatbot, the method comprising: receiving user information corresponding to a user; resolving the received user information to a unique identifier (UID) for the user; obtaining, from one or more data sources, comprehensive skills-or-knowledge based data about the user that matches the UID; generating a plurality of skills-or-knowledge based authentication questions based on the comprehensive skills-or-knowledge based data; outputting for display on a user device, one or more of the plurality of the skills-or-knowledge based authentication questions; receiving, in a natural language, one or more user answers corresponding to the one or more of the plurality of the skills-or-knowledge based authentication questions; utilizing a large language model (LLM) to evaluate the one or more user answer against a known ground truth based on the comprehensive skills-or-knowledge based data; and responsive to determining that a success count corresponding to a number of factually correct user answers corresponding to the one or more of the plurality of the skills-or-knowledge based authentication questions matches a pre-defined threshold number of matches, authenticating the user for access to the service. ### Claim 9 (depends on 8) The computer-implemented method of claim 8, further comprising: gathering user device data and behavioral data responsive to the user accessing an authentication page; and generating a digital profile associated with the user. ### Claim 10 (depends on 8) The computer-implemented method of claim 8, wherein generating the plurality of the skills-or-knowledge based authentication questions is further based on a pre-defined configuration. ### Claim 11 (depends on 8) The computer-implemented method of claim 8, wherein: If a user answer is evaluated as factually correct, outputting for display on the user device, a next question and incrementing the success count; if a user answer is evaluated as factually incorrect, outputting for display on the user device, a next question without incrementing the success count; if a user answer is factually correct but incomplete, outputting for display on the user device, additional instructions to indicate missing information without incrementing the success count; and if a user answer is incorrect and incomplete, outputting for display on the user device, additional instructions to indicate missing information without incrementing the success count. ### Claim 12 (depends on 8) The computer-implemented method of claim 8, wherein the one or more data sources comprise one or more of: court records, birth certificate records, census bureau records, county records, property records, real estate records, court records, business records, school records, education records, webpages, professional license records, historical records, social media records, marketing records, publications, private records, media records, background records, proprietary data, and derived data records. ### Claim 13 (depends on 8) The computer-implemented method of claim 8, further comprising gathering user device and behavioral data responsive to the user accessing an authentication page to generate a digital profile associated with the user. ### Claim 14 (depends on 8) The computer-implemented method of claim 8, wherein the one or more skills-or-knowledge based authentication questions includes a photograph of an apparatus associated with the comprehensive skills-or-knowledge based data about the user that matches the UID. ### Claim 15 (depends on 8) The computer-implemented method of claim 8, wherein the one or more of the plurality of the skills-or-knowledge based authentication questions includes one or more open-ended questions. ### Claim 16 A system for user authentication via a chatbot, the system comprising: a data repository configured for storing user identification information; a user interface configured for displaying authentication questions and receiving user answers from a user; at least one memory for storing data and computer-executable instructions; and at least one processor configured to access the at least one memory and further configured to execute the computer-executable instructions that cause the at least one processor to: receive user information corresponding to a user; resolve the received user information to a unique identifier (UID) for the user; obtain, from one or more data sources, comprehensive data about the user that matches the UID; generate a plurality of authentication questions based on the comprehensive data; output for display on a user device, one or more of the plurality of authentication questions; receive, in a natural language, one or more user answers corresponding to the one or more of the plurality of authentication questions; utilize a large language model (LLM) to evaluate the user answer against a known ground truth based on the comprehensive data; output for display on the user device, an indication corresponding to the evaluation; and authenticate the user for access to a service responsive to determining that a success count corresponding to a number of factually correct user answers corresponding to the one or more of the plurality of authentication questions matches a pre-defined threshold number of matches. ### Claim 17 (depends on 16) The system of claim 16, wherein the plurality of authentication questions comprises one or more of open-ended questions, skills-based questions, and knowledge-based questions. ### Claim 18 (depends on 16) The system of claim 16, wherein the system is further configured to: gather user device data and behavioral data responsive to the user accessing an authentication page; and generate a digital profile associated with the user. ### Claim 19 (depends on 16) The system of claim 16, wherein the authentication questions are further generated based on a pre-defined configuration. ### Claim 20 (depends on 16) The system of claim 16, wherein the system is further configured to output for display on the user device, a next question and increment the success count when a user answer is evaluated as factually correct; output for display on the user device, a next question without incrementing the success count when a user answer is evaluated as factually incorrect; output for display on the user device, additional instructions to indicate missing information without incrementing the success count when a user answer is factually correct but incomplete; and output for display on the user device, additional instructions to indicate missing information without incrementing the success count when a user answer is incorrect and incomplete.