Web browser privacy: Popular desktop web browsers ability to continuously spoof their fingerprint

Abstract: Background. Web tracking is a constant threat to our privacy when browsing the web. There exist multiple methods of tracking, but browser fingerprinting is more elusive and difficult to control. Browser fingerprinting works by a website collecting all kinds of browser and system information on visiting clients and then combining those into one set of information that can uniquely identify users. Objectives. In this thesis, we tested three of today's most used web browsers for the desktop platform to determine their ability to utilize one type of countermeasure, attribute spoofing. We aimed at determining how the browsers perform in two cases. The first case is when running with a default configuration. The second case is when the attribute spoofing is improved with the help of both altered settings and installed extensions. We also aimed at determining if the choice of browser matters in this aspect. Methods. The method for determining these goals was to conduct an experiment to collect 60 fingerprints from each browser and determine the effectiveness of the attribute spoofing via a weight-based system. We also used statistics to see the value range for spoofed attributes and to determine if any browser restart is required for certain spoofing to occur. Results. Our results show little to no attribute spoofing when browsers run in their default configuration. However, significant improvements were made through anti-fingerprint extensions. Conclusions. Our conclusion is, if the tested browsers' do not utilize any other type of countermeasure than attribute spoofing, using browsers at their default configuration can result in a user being alarmingly vulnerable to browser fingerprinting. Installing extensions aimed at improving our protection is therefore advised.