Abstract: Embodiments of the invention are directed to an automated account restoration system. In some embodiments, the system determines a state of an account based on a likelihood that the account has been compromised. If the account is determined to be in a low-risk state, then upon an successful login to that account, a verification cookie may be generated which is unique to a user device used to access the account. If the account is determined to be in a high-risk state, then system may prevent access to the account except by user devices that include a valid verification cookie.

Abstract: A method and system for remediating vulnerable code libraries, including open source libraries, in a software application. An application, that uses code libraries, and information regarding known library vulnerabilities are received, then it is determined if one or more libraries in the application are vulnerable based upon the information. For each of the one or more vulnerable libraries, a library version that minimizes risk is determined. The determined library version is incorporated into the application to form a test application, and an application test is performed on the test application. If an application test score on the test application is below a predetermined threshold, the determined library version is incorporated into a final application precursor. A final application can be determined from the final application precursor for each of the one or more vulnerable libraries.

Abstract: A method and system for remediating vulnerable code libraries, including open source libraries, in a software application are disclosed. An application that uses code libraries and information regarding known library vulnerabilities is received, then it can be determined if one or more libraries in the application is vulnerable based upon the information. For each of the one or more vulnerable libraries a library version that minimizes risk is determined. The determined library version is incorporated into the application to form a test application, and an application test is performed on the test application. If the application test is below a predetermined threshold, the determined library version is incorporated into a final application precursor. A final application can be determined from the final application precursor for each vulnerable library.

Abstract: A method and system for remediating vulnerable code libraries, including open source libraries, in a software application. An application that uses code libraries and information regarding known library vulnerabilities are received, then it identifies one or more libraries in the application that are vulnerable based upon the information. For each of the one or more vulnerable libraries, a library version that minimizes risk is determined. The determined library version is incorporated into the application to form a test application, and an application test is performed on the test application. If an application test score on the test application is below a predetermined threshold, the determined library version is incorporated into a final application precursor. A final application can be determined from the final application precursor for each of the one or more vulnerable libraries.

Abstract: Techniques for securely generating and using a “fingerprint” for authentication. A server computer receives a first data set from a user device (including a first fuzzy hash of first user data on the user device). The server computer generates a first fingerprint value based on the first data set. The server computer detects an event corresponding to a user in association with the user device. The server computer identifies a baseline fingerprint value (generated based on a baseline fuzzy hash of user data on the user device). The server computer compares the first fingerprint value to the baseline fingerprint value to generate a similarity score. The server computer may determine that the similarity score exceeds a threshold value but does not represent an exact match, and, based on the similarity score, authenticate the user and update the baseline fingerprint value based on the first fingerprint value.

Abstract: Embodiments of the invention are directed to an automated account restoration system. In some embodiments, the system determines a state of an account based on a likelihood that the account has been compromised. If the account is determined to be in a low-risk state, then upon an successful login to that account, a verification cookie may be generated which is unique to a user device used to access the account. If the account is determined to be in a high-risk state, then system may prevent access to the account except by user devices that include a valid verification cookie.

Abstract: A method and system for remediating vulnerable code libraries, including open source libraries, in a software application are disclosed. The method includes receiving an application that uses code libraries and information regarding known library vulnerabilities, then determining if one or more libraries in the application is vulnerable based upon the information. For each of the one or more vulnerable libraries a library version that minimizes risk is determined. The determined library version is incorporated into the application to form a test application, and an application test is performed on the test application. If the application test is below a predetermined threshold, the determined library version is incorporated into a final application precursor. A final application can be determined from the final application precursor for each vulnerable library.

Abstract: Techniques for securely generating and using a “fingerprint” for authentication. A server computer receives a first data set from a user device (including a first fuzzy hash of first user data on the user device). The server computer generates a first fingerprint value based on the first data set. The server computer detects an event corresponding to a user in association with the user device. The server computer identifies a baseline fingerprint value (generated based on a baseline fuzzy hash of user data on the user device). The server computer compares the first fingerprint value to the baseline fingerprint value to generate a similarity score. The server computer may determine that the similarity score exceeds a threshold value but does not represent an exact match, and, based on the similarity score, authenticate the user and update the baseline fingerprint value based on the first fingerprint value.