Safe Browsing Services: to Track, Censor and Protect

Thomas Gerbet (1) , Amrit Kumar (2) , Cédric Lauradoux (2)

1 UJF - Université Joseph Fourier - Grenoble 1
2 PRIVATICS - Privacy Models, Architectures and Tools for the Information Society

Users often inadvertently click malicious phishing or malware website links, and in doing so they sacrifice secret information and sometimes even fully compromise their devices. The purveyors of these sites are highly motivated and hence construct intelligently scripted URLs to remain inconspicuous over the Internet. In light of the ever increasing number of such URLs, new ingenious strategies have been invented to detect them and inform the end user when he is tempted to access such a link. At the forefront lies the \textit{Safe Browsing} technique, which queries a blacklist to identify unsafe websites and notify users and webmasters allowing them to protect themselves from harm. In this paper, we show how Safe Browsing services work and how they damage the privacy of their users. We study the privacy of high impact Safe Browsing services including Google/ and Yandex to understand how much information they can collect. We hereby warn their users that these services can be easily used for surveillance to track specific classes of individuals or to censor certain websites. Google and Yandex are currently not accountable for the entries included in their blacklist. Until a major advance in the field of private information retrieval, Safe Browsing can not be made private and the current services can not be trusted. We provide evidences that \yandex/ has all probes in place to track the users of targeted websites.

Mots clés: Malware; phishing; surveillance

Domaines: Cryptographie et sécurité [cs.CR]; Informatique et langage [cs.CL]

Thomas Gerbet, Amrit Kumar, Cédric Lauradoux. Safe Browsing Services: to Track, Censor and Protect. [Research Report] RR-8686, INRIA. 2015. ⟨hal-01120186v2⟩