Crowdsourced, self-learning security system through smart feedback loops

A computing system comprising:

• one or more processors;
• and
• one or more computer-readable media having stored thereon instructions that are executable by the one or more processors to configure the computer system to implement a method of improving computer system security, including instructions that are executable to configure the computer system to implement the method by performing at least the following:
• obtain one or more user risk reports identifying a plurality of users and a corresponding user risk profile for each user in the plurality of users;
• generate an interface display that identifies the plurality of users in the one or more user risk reports and their corresponding risk profile that specifies a risk level and a risk state, along with one or more selectable display objects that, when selected, are operable for confirming or changing the specified risk state of each user;
• receive user input selecting one or more of the selectable display objects for one or more corresponding user risk profiles and, responsively, confirming or changing the risk state of the one or more corresponding user risk profiles;
• generate a label data file that contains a listing of the corresponding user risk profiles and that specifies each corresponding user risk profile in the listing as either a false positive, false negative, true positive or true negative, based on the one or more user risk reports and the user input;
• and
• improve or modify a machine learning tool with the label data file and which machine learning tool is used to generate or modify a risk profiles report that is used by a risk assessment engine to generate user risk scores, and wherein the user risk scores are used to generate or update the one or more user risk reports.