It's (DOM) Clobbering Time: Attack Techniques, Prevalence, and Defenses

Summary

DOM Clobbering is a type of code-less injection attack where attackers insert a piece of non-script, seemingly benign HTML markup into a webpage and transform it to executable code by exploiting the unforeseen interactions between JavaScript code and the runtime environment. The attack techniques, browser behaviours, and vulnerable code patterns that enable DOM Clobbering has not been studied yet, and in this paper, we undertake one of the first evaluations of the state of DOM Clobbering on the Web platform. Starting with a comprehensive survey of existing literature and dynamic analysis of 19 different mobile and desktop browsers, we systematize DOM Clobbering attacks, uncovering 31.4K distinct markups that use five different techniques to unexpectedly overwrite JavaScript variables in at least one browser. Then, we use our systematization to identify and characterize program instructions that can be overwritten by DOM Clobbering, and use it to present TheThing, an automated system that detects clobberable data flows to security-sensitive instructions. We instantiate TheThing on the top of the Tranco top 5K sites, quantifying the prevalence and impact of DOM Clobbering in the wild. Our evaluation uncovers that DOM Clobbering vulnerabilities are ubiquitous, with a total of 9,467 vulnerable data flows across 491 affected sites, making it possible to mount arbitrary code execution, open redirections, or client-side request forgery attacks also against popular websites such as Fandom, Trello, Vimeo, TripAdvisor, WikiBooks and GitHub, that were not exploitable through the traditional attack vectors. Finally, in this paper, we also evaluate the robustness of the existing countermeasures, such as HTML sanitizers and Content Security Policy, against DOM Clobbering.