Quantifying Web Adblocker Privacy

Web advertisements, an integral part of today's web browsing experience, financially support countless websites. Meaningful advertisements, however, require behavioral targeting, user tracking and profile fingerprinting that raise serious privacy concerns. To counter privacy issues and enhance usability, adblockers emerged as a popular way to filter web requests that do not serve the website's main content. Despite their popularity, little work has focused on quantifying the privacy provisions of adblockers. In this paper, we develop a quantitative framework to compare the privacy provisions of adblockers objectively. For our methodology, we introduce several privacy metrics that capture not only the technical web architecture but also the underlying corporate institutions of the problem across time and geography. Using our framework, we quantify the web privacy implications of 12 ad-blocking software combinations and browser settings on 1000 websites on a daily basis over a timespan of three weeks (a total of 252'000 crawls). Our results highlight a significant difference among adblockers regarding filtering performance, in particular, affected by the applied configurations. Our experimental results confirm that our framework provides consistent results and hence can be used as a quantitative methodology to assess other configurations and adblockers further.