Artificial intelligence privacy protection for cybersecurity analysis

Abstract

A privacy protection component can automatically comply with a set of privacy requirements when displaying input data. An ingestion module collects input data describing network activity executed by a network entity. A clustering module identifies data fields with data values within the input data as data identifiable to the network entity using machine-learning models trained on known data fields and their data. The clustering module also clusters the data values with other data values having similar characteristics using machine-learning models to infer a privacy level associated with each data field. The privacy level is utilized to indicate whether a data value in that data field should be anonymized. A permission module determines a privacy status of that data field by comparing the privacy level from the clustering module to a permission threshold. An aliasing module applies an alias transform to the data value of that data field with a privacy alias to anonymize that data value in that data field. A user interface module displays the input data to a system user with the privacy alias from the aliasing module substituted for the data value for that data field.