Detecting malicious code existing in internet advertisements by ongoing sandbox monitoring

Abstract

There are disclosed devices, system and methods for detecting malicious code existing in an internet advertisement (ad) requested by a published webpage viewed by a user. First, receipt of malicious code of the ad is detected, where that code may be malicious code that causes a browser unwanted action without user action. If the internet ad is an SCR type document, the malicious code may be wrapped in a java script (JS) closure to detect an unwanted action requested by the malicious code. The malicious code is executed a browser sandbox that activates the unwanted action, that displays execution of the internet ad and that allows execution of the unwanted action. When a security error resulting from the unwanted action is detected, executing the malicious code in the browser sandbox is discontinued, displaying of the internet ad on the display is discontinued, and execution of the unwanted action is stopped.