Internet access control based on external third-party data

A method of controlling access to the Internet, the method comprising:

• retrieving, over the Internet, event data of scheduled events of a user from a third-party scheduling calendar application;
• sub-dividing each day of other event data in the third-party scheduling calendar application into a plurality of time windows;
• generating training data by, for each time window in the plurality of time windows, marking an event within the time window as a positive sample indicative of prohibited Internet access when an access control policy of a plurality of pre-existing control policies is in effect during the time window and marking the event within the time window as a negative sample indicative of permitted Internet access when none of the plurality of pre-existing control policies is in effect during the time window;
• training a machine learning model using the training data;
• using the machine learning model to determine if the event data are indicative of permitted or prohibited Internet access;
• creating an access control policy based on determination made by the machine learning model on the event data;
• enforcing the access control policy in an Internet access control device;
• detecting, by the Internet access control device, an Internet access of the user to connect to a target server;
• blocking, by the Internet access control device, the Internet access when the access control policy indicates that the Internet access is a prohibited Internet access;
• and
• allowing, by the Internet access control device, the Internet access when the access control policy indicates that the Internet access is a permitted Internet access.