Microsoft Technology Licensing, LLC

Rajat Luthra

Maria Puertas Calvo

Sayed Hassan Abdelaziz

Laurentiu Bogdan Cristofor

Lee Walker

Zachary Michael Edwards

Colin Trent

Poulomi Bandyopadhyay

Authentication request notifications are selectively suppressed, reducing notification fatigue and susceptibility to social engineering attacks. Authentication request notifications may be suppressed by not presenting a push notification on the user's phone. The authentication request may still be accessed and approved by manually opening the authenticator app. Notifications may be suppressed based on an estimation that the person attempting to login is not who they say they are. This estimation may be based on applying heuristics and/or machine learning models to the context of the login attempt, such as the IP address that originated the login request, time of day, recent user actions, patterns of previous logins, etc. One heuristic determines that the user has repeatedly ignored notifications caused by a particular IP address. Machine learning models generate a risk score from the login context, and notifications may be suppressed if the risk score exceeds a threshold.

Selectively and intelligently displaying authentication notifications to protect users

Systems are provided for improving computer security systems that are based on user risk scores. These systems can be used to improve both the accuracy and usability of the user risk scores by applying multiple tiers of machine learning to different the user risk profile components used to generate the user risk scores and in such a manner as to dynamically generate and modify the corresponding user risk scores.

Supervised learning system for identity compromise risk computation

Systems are provided for utilizing crowdsourcing and machine learning to improve computer system security processes associated with user risk profiles and sign-in profiles. Risk profiles of known users and logged sign-ins are confirmed by user input as either safe or compromised. This input is used as crowdsourced feedback to generate label data for training/refining machine learning algorithms used to generate corresponding risky profile reports. The risky profile reports are used to provide updated assessments and initial assessments of known users and logged sign-ins, as well as newly discovered users and new sign-in attempts, respectively. These assessments are further confirmed or modified to further update the machine learning and risky profile reports.

Crowdsourced, self-learning security system through smart feedback loops

Growing community of inventors

Redmond, WA, United States of America

Rajat Luthra