User Profiles: The Achilles’ Heel of Web Browsers
Dolière Francis Somé, Moaz Airan, Zakir Durumeric, Cristian-Alexandru Staicu · 2025
Read User Profiles: The Achilles’ Heel of Web Browsers by Dolière Francis Somé, Moaz Airan, Zakir Durumeric, Cristian-Alexandru Staicu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-17692.txt.
Introduction
Web browsers guard our online lives, yet the files they leave on disk get little security scrutiny. Dolière Francis Somé and co-authors present the first systematic study of browser profiles, the folders holding cookies, extensions, saved passwords, and trust settings. Except for the Tor Browser, they find, modern browsers store this material with weak or missing protection. Password and cookie encryption can be bypassed, and a planted root certificate can quietly undo HTTPS protection altogether.
Worse, a Chromium feature that lets websites touch the file system can mount these attacks from an ordinary web page. Case studies show planted extensions, hijacked secure traffic, and camera access.
Catalog record
| CREATOR | Dolière Francis Somé, Moaz Airan, Zakir Durumeric, Cristian-Alexandru Staicu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 92 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-17692
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
User Profiles: The Achilles’ Heel of Web Browsers Dolière Francis Somé, Moaz Airan, Zakir Durumeric, Cristian-Alexandru Staicu Abstract . Web browsers provide the security foundation for our online experiences. Significant research has been done into the security of browsers themselves, but relatively little investigation has been done into how they interact with the operating system or the file system. In this work, we provide the first systematic security study of browser profiles, the on-disk persistence layer of browsers, used for storing everything from users’ authentication cookies and browser extensions to certificate trust decisions and device permissions. We show that, except for the Tor Browser, all modern browsers store sensitive data in home directories with little to no integrity or confidentiality controls. We show that security measures like password and cookie encryption can be easily bypassed. In addition, HTTPS can be sidestepped entirely by deploying malicious root certificates within users’ browser profiles. The Public Key Infrastructure (PKI), the backbone of the secure Web—HTTPS—can be fully bypassed with the deployment of custom potentially malicious root certificates. More worryingly, we show how these powerful attacks can be fully mounted directly from web browsers themselves, through the File System Access API, a recent feature added by Chromium browsers that enables a website to directly manipulate a user’s file system via JavaScript. In a series of case studies, we demonstrate how an attacker can install malicious browser extensions, inject additional root certificates, hijack HTTPS traffic, and enable websites to access hardware devices like the camera and GPS. Based on our findings, we argue that researchers and browser vendors need to develop and deploy more secure mechanisms for protecting users’ browser data against file system attackers. Volume: YYYY X DOI: XXXXXXX.XXXXXXX 1. Introduction To mediate users’ interactions with the web, browsers have deployed dozens of security mechanisms. This includes sandboxing untrusted Javascript code ( SOP, 2024 ) , hardening browser media renderers to isolate websites from one another ( COR, 2024 ) , and fine-grained permissions mechanisms for accessing device hardware ( CSP, 2024 ) . The academic community published hundreds of papers studying the limits ( Luo et al., 2019 ; Nomoto et al., 2023 ; Snyder et al., 2023 ) or testing ( Bernardo et al., 2024 ; Shou et al., 2021 ; Zhou et al., 2022 ; Brown et al., 2020 ) defenses for web browsers, leading to significant improvements. Despite the plethora of work analyzing how browsers internally operate and how they can defend against attacks targetting visited websites during user’s browsing sessions, there has been relatively little work investigating how resilient browsers are against attackers that have access to the file system . Traditionally, such attackers were considered too strong and deemed out of scope in browsers’ threat models. However, recent developments urge us to consider this threat model for modern browsers. Below, we list three essential supporting arguments. First, the advent of the File System Access (FSA) API in Chromium browsers allows web attackers to access the file system directly. Oz et al. ( Oz et al., 2023 ) show that this API can be misused for ransomware attacks, but they argue that it cannot be abused for stealing browsers’ data because of the blocklist mandated by the FSA standard ( Sullivan, 2024 ) . However, as noted by the Chromium team ( Developer, 2024 ; Team, 2024 ) , this blocklist is not a silver bullet but a pragmatic defense against blatant misuse. In this work, we demonstrate that remote attackers can leverage the FSA API to compromise data stored by browsers on the disk. Second, the landscape of unwanted software ( Urban et al., 2018 ) has changed significantly in recent years with the advent of prevalent supply chain attacks ( Ohm et al., 2020 ) . Traditionally, browser vendors often consider defenses against malware present on the machine out of scope ( Team, 2024 ) , arguing that such attackers can mount more powerful attacks, such as altering the browser’s binaries. However, there are several recent reports of open-source packages stealing information persisted by browsers on the disk: passwords ( Vaas, 2021 ) , session cookies ( Npm packages spread ’Bladeroid’ crypto stealer, 2024 ) , entire browser profiles ( “express-dompurify” npm Package Steals Browser and Data, 2024 ) , extensions’ data ( Data, 2024 ) . These malicious packages have complete power over the victim’s machine, but they specifically target these browser artifacts, allowing them to carry out web attacks ( Kiat and Gupta, 2024 ; Shen, 2021 ) . Third, we already see initial attempts by the vendors to deploy defenses against file system attackers. For example, as discussed in Section 2 , Chromium-based browsers and Firefox encrypt user passwords before storing them on the disk. Picazo-Sanchez et al. ( Picazo-Sanchez and Sabelfeld, 2020 ) also discuss how a simple HMAC mechanism in Chromium browsers is implemented to protect preference files and extensions’ integrity. Our paper shows that these mechanisms are insufficient against file system attackers. Notably, in concurrent and independent work, browser vendors are exploring more principled solutions for protecting their persistence layer, using device-bound session credentials (DBSC) ( cookie theft using device bound sessions, 2024 ) that aims to prevent cookie theft or application-bound encryption on Windows ( the security of Chrome cookies on Windows, 2024 ) . While these proposals show that browser vendors are taking file system attackers seriously, they are in an early phase and, thus, still to be widely adopted by all vendors and operating systems. Considering the vendors’ interest and a high potential for securing and compromising browser profiles, we systematically analyze the data stored by modern browsers on the disk and ways of compromising it. We answer the following research questions: RQ1: What data do browsers store in user-accessible file system locations, and what security mechanisms are implemented to protect it against external manipulation? (Section 2 ) Web browsers store user-specific data and configuration files in browser profile folders . When a user first launches a browser, it creates and stores the user profile data in well-known locations depending on the operating system, browser, and configurations ( Chr, 2024b ; Fir, 2024b ) . A user’s profile consists of an on-disk collection of databases, configuration, and preferences files that the browser writes to preserve the state of the browser. Depending on the browser, this profile contains data like installed extensions, stored usernames and passwords, cookies, bookmarks, custom root certificate authorities, proxy preferences, and permissions for accessing physical devices such as cameras or microphones. A subset of this data, including cookies and passwords, are encrypted on disk. Another subset, including browser extensions, has integrity checks. Other configurations, like device access and trusted certificate authorities, have no protections. Unlike many application configuration files, we emphasize that a browser’s configuration can fundamentally alter its security-relevant behavior. We find that any application that can write to a user’s home directory (or equivalent) can easily install code into a browser, proxy traffic to attacker websites, and change device permissions simply by changing on-disk configuration files. RQ2: What attacks can be mounted against browsers, users, and websites by tampering with profile folders? (Sections 4 and 5 ) We show that several attacks can