Identification card based multi-factor authentication

Patent metadata

Field	Value
Publication number	US-12476965-B2
Application number	US-202418435351-A
Country	US
Kind code	B2
Filing date	Feb 7, 2024
Priority date	Feb 7, 2024
Publication date	Nov 18, 2025
Grant date	Nov 18, 2025

Abstract

Systems and methods for improvements to multifactor authentication (MFA) techniques are described herein. The improved and enhanced MFA techniques can leverage user interaction data to predict an entity-issued user-specific identification document that an authorized user of an authorized user account is most likely to carry on their person. Upon a request to access an authorized user account being received, the improved and enhanced MFA techniques can request that the user capture, in real-time, an image depicting the predicted identification document. Using additional artificial intelligence, a determination may be made as to whether the image depicts a real instance of the entity-issued user-specific identification document. This determination may result in access to the authorized user account being provided or denied.

First claim

What is claimed is:

1 . A system for providing cryptographically-secure access to an authorized user account, the system comprising:
one or more processors programmed using computer program instructions to:
receive a request to access an authorized user account of a user;
retrieve user interaction data associated with the authorized user account, wherein the user interaction data comprises prior interactions of the user with a set of entities;
obtain a list of entities that are known to issue physical user identification cards, wherein the list of entities includes at least one of the set of entities;
predict, using a trained artificial intelligence model, based on the request, the user interaction data, and the list of entities, a physical user identification card associated with an entity from the set of entities that the user is predicted to currently carry on their person;
identify, based on the authorized user account, a user device of the user to provide authentication instructions;
send the authentication instructions to the user device, the authentication instructions comprising instructions to capture an image of the physical user identification card using the user device;
calculate, using a trained computer vision model, a user identification card score indicating a likelihood that the image depicts a real instance of the physical user identification card of the user for which the user is predicted to currently carry on their person;
provide, based on the user identification card score being greater than or equal to a threshold confidence score indicating that the image depicts the real instance of the physical user identification card of the user, access to the authorized user account of the user.

2 . The system of claim 1 , wherein the user identification card score is calculated based on at least one of:
a facial recognition confidence score determined using a facial recognition component of the trained computer vision model indicating a likelihood that at least a portion of the image depicts a face of the user; a character recognition confidence score determined using a character recognition component of the trained computer vision model indicating a likelihood that at least a portion of the image depicts one or more alphanumeric characters known to present on the real instance of the physical user identification card; or a logo recognition confidence score determined using a logo recognition component of the trained computer vision model indicating a likelihood that at least a portion of the image depicts at least a portion of a logo known to be present on the real instance of the physical user identification card.

3 . The system of claim 2 , wherein the user identification card score is determined based on a weighted combination of the at least one of the facial recognition confidence score, the character recognition confidence score, and the logo recognition confidence score.

4 . The system of claim 1 , wherein the physical user identification card associated being predicted comprises the one or more processors being programmed to:
filter the list of entities based on the set of entities to identify (i) a subset of entities with whom the user has had a prior interaction and (ii) a subset of the prior interactions associated with the subset of entities; and determine, based on the subset of the prior interactions, a frequency that the user interacted with each of the subset of entities, wherein the frequency associated with each entity of the subset of entities is input to the trained artificial intelligence model to determine how likely the user carries that entities corresponding physical user identification card.

5 . A method for providing cryptographically-secure access to an authorized user account, the method being implemented by one or more processors, the method comprising:
receiving a request to access an authorized user account of a user; determining, using a trained artificial intelligence model, an entity-issued user-specific identification document predicted to be currently possessed by the user based on the request, prior interactions of the user with a set of entities, and a list of entities known to issue tangible identification documents; causing a user device associated with the user to capture an image of the entity-issued user-specific identification document using the user device; determining, using a trained computer vision model, a user identification card score based on the image; generating, based on the user identification card score and a threshold condition, an authorized access result indicating that access to the authorized user account has been granted or denied.

6 . The method of claim 5 , wherein determining the user identification card score comprises:
determining, using the trained computer vision model, a facial recognition confidence score, wherein the user identification card score is based on the facial recognition confidence score.

7 . The method of claim 5 , wherein determining the user identification card score comprises:
determining, using the trained computer vision model, a character recognition confidence score, wherein the user identification card score is based on the character recognition confidence score.

8 . The method of claim 5 , wherein determining the user identification card score comprises:
determining, using the trained computer vision model, a logo recognition confidence score, wherein the user identification card score is based on the logo recognition confidence score.

9 . The method of claim 5 , wherein determining the user identification card score comprises:
determining, using the trained computer vision model, a weighted combination of a facial recognition confidence score, a character recognition confidence score, and a logo recognition confidence score, the user identification card score comprising the weighted combination.

10 . The method of claim 5 , further comprising:
determining, based on the request, a request risk score indicating how much risk the request places on the authorized user account; and determining, based on the request risk score satisfying a request risk threshold score, that multi-factor authentication is needed to allow access to the authorized user account of the user, wherein user interaction data comprising the prior interactions of the user is retrieved based on the request risk score satisfying the request risk threshold score.

11 . The method of claim 5 , wherein causing comprises:
generating or obtaining user interface data for rendering a user interface on the user device, wherein the user interface comprises instructions for accessing an image capture component of the user device and capturing the image of a physical user identification card.

12 . The method of claim 5 , wherein determining the entity-issued user-specific identification document comprises:
generating, using the trained artificial intelligence model, a set of entity-likelihood scores respectively corresponding to the set of entities, wherein each entity has an entity-likelihood score representing a likelihood that the user carries the entity-issued user-specific identification document for the entity; and selecting the entity-issued user-specific identification document based on the entity-likelihood score associated therewith.

13 . The method of claim 5 , further comprising:
retrieving, based on the request, user interaction data associated with the authorized user account, wherein the user interaction data comprises the prior interactions of the user with the set of entities; obtaining, based on the request, the l

Assignees

• Capital One Services Llc

Inventors

• Edwards Joshua
• Bowers Justin Anthony
• Cody Lea
• Teramoto Elaina
• Maiman Tyler
• Sweeney Mary
• Neighbour Erik
• Kavi Anantha Swarna

Classifications

• H04L63/0861 Primary

  using biometrical features, e.g. fingerprint, retina-scan (cryptographic mechanisms or cryptographic arrangements for entity authentication using biological data H04L9/3231) · CPC title

• G06N3/084

  Backpropagation, e.g. using gradient descent · CPC title

• H04L63/0853 Primary

  using an additional device, e.g. smartcard, SIM or a different communication terminal (cryptographic mechanisms or cryptographic arrangements for entity authentication involving additional secure or trusted devices H04L9/3234) · CPC title