Secure generation of authentication datasets from network activity
Disclosed are example methods, systems, and devices that allow for secure generation of authentication datasets from network activity. The techniques include accessing secured data sources to generate a first dataset of secured data, and extracting information from one or more unsecured data sources to generate a second dataset comprising a second dataset. A third set of data elements can be generated from the first and second datasets, and may be utilized to authenticate credentials that can be utilized to access secured data via a network. The techniques can transmit indications that credentials are invalid if the credentials fail to satisfy aspects of the third dataset.
1. A method implemented by a first computing system comprising a server, the method comprising:
generating, by the first computing system, a first dataset comprising security elements corresponding to one or more accounts administered by the first computing system, the one or more accounts comprising a first account, the security elements comprising one or more credentials granting access to the first account via the server;
obtaining, by the first computing system, network access to one or more secured data sources, the one or more secured data sources comprising a second computing system that is unfederated with respect to the first computing system;
accessing, by the first computing system, the one or more secured data sources to generate a second dataset based on data in the second computing system;
extracting, by the first computing system, from one or more unsecured data sources comprising a third computing system that is unfederated with respect to the first computing system, a third dataset based on data in the third computing system;
applying, by the first computing system, a machine-learning model to the second dataset and the third dataset to generate a fourth dataset comprising recurrent data elements, the machine-learning model comprising a natural language processing (NLP) model;
determining, by the first computing system, based on the first dataset and the fourth dataset, that a network security threat corresponding to the first account exceeds a threshold;
receiving, by the server, via a login page of a web portal, a request to access the first account;
transmitting, by the server, an indication that replacement credentials are required to access the first account;
receiving, by the server via the web portal, one or more replacement credentials, and determining that the one or more replacement credentials are acceptable based on the fourth dataset;
registering, by the first computing system, the one or more replacement credentials to the first account; and
granting, by the server, via the web portal, access to the first account based on the one or more replacement credentials.
2. The method of claim 1 , further comprising:
providing, by the server, the web portal with a registration page for registering for a digital security service; and
receiving, via the registration page, a request to register for the digital security service.
3. The method of claim 2 , wherein the network access to the one or more secured data sources is obtained based on inputs received via the registration page of the web portal.
4. The method of claim 1 , wherein the one or more credentials granting access to the first account comprises a passcode.
5. The method of claim 1 , wherein the one or more credentials granting access to the first account comprises a response to a security question corresponding to the first account.
6. The method of claim 1 , wherein the second dataset comprises one or more documents.
7. The method of claim 1 , wherein the second dataset comprises one or more audio recordings.
8. The method of claim 1 , wherein the second dataset comprises one or more video recordings.
9. The method of claim 1 , wherein extracting the third dataset comprises web scraping data from one or more web pages.
10. The method of claim 1 , wherein the third dataset indicates a data breach by a compromised computing system.
11. The method of claim 10 , further comprising determining, by the first computing system, that the compromised computing system likely stored data on a user of the first account.
12. The method of claim 11 , wherein determining that the compromised computing system likely stored data on the user is based at least partly on account activity in the first account.
13. The method of claim 12 , wherein the account activity identifies an electronic transaction with an entity corresponding to the compromised computing system.
14. The method of claim 1 , wherein accessing, by the first computing system, the one or more secured data sources comprises periodically retrieving data from the one or more secured data sources.
15. The method of claim 14 , wherein periodically retrieving data from the one or more secured data sources comprises transmitting, by the first computing system, a plurality of API calls to the second computing system.
16. The method of claim 15 , wherein one or more API calls are transmitted to the second computing system on a weekly basis.
17. The method of claim 16 , wherein the one or more API calls comprise a security token identifying the first computing system.
18. The method of claim 16 , wherein the one or more API calls comprise a security token identifying the first account.
19. A first computing system comprising a server, the first computing system comprising one or more processors configured to:
generate a first dataset comprising security elements corresponding to one or more accounts administered by the first computing system, the one or more accounts comprising a first account, the security elements comprising one or more credentials granting access to the first account via the server;
obtain network access to one or more secured data sources, the one or more secured data sources comprising a second computing system that is unfederated with respect to the first computing system;
access the one or more secured data sources to generate a second dataset based on data in the second computing system;
extract, from one or more unsecured data sources comprising a third computing system that is unfederated with respect to the first computing system, a third dataset based on data in the third computing system;
applying a machine-learning model to the second dataset and the third dataset to generate a fourth dataset comprising recurrent data elements, the machine-learning model comprising a natural language processing (NLP) model;
determine, based on the first dataset and the fourth dataset, that a network security threat corresponding to the first account exceeds a threshold;
receive, by the server, via a login page of a web portal, a request to access the first account;
transmit, by the server, an indication that replacement credentials are required to access the first account;
receive, by the server via the web portal, one or more replacement credentials, and determine that the one or more replacement credentials are acceptable based on the fourth dataset;
register the one or more replacement credentials to the first account; and
grant, by the server, via the web portal, access to the first account based on the one or more replacement credentials.
20. The computing system of claim 19 , the one or more processors further configured to:
provide, by the server, the web portal with a registration page for registering for a digital security service; and
receive, via the registration page, a request to register for the digital security service, wherein the network access to the one or more secured data sources is obtained based on inputs received via the registration page of the web portal.