Our project aims to integrate AI into the popular ad blocker uBlock Origin as a means to counter phishing attempts.
Currently, modern ad blockers rely on static blacklists and predefined sets of rules, which are weak points in adapting to new and evolving phishing threats and malicious web activities.
By creating a fast and light model which can assess whether a site is malicious or not according to relevant features of the URL and site's content, it'll be able to smartly and swiftly decide whether the site is to be trusted or not and pass that feedback to the user.