Poster: When Brave Hurts Privacy: Why Too Many Choices do More Harm Than Good Anna Lena Fehlhaber, Yasemin Acar, and Sascha Fahl Leibniz University Hannover Email: {lastname}@sec.uni-hannover.de Marco Gutfleisch, Daniel Theis, and Florian Wallk¨ otter Ruhr University Bochum Email: {firstname.lastname}@rub.de Abstract—Private browsing modes offer users various privacy features. However, users have misconceptions about what these privacy features can and cannot accomplish. They generally expect local and network protection, while private browsing modes only offer local protection by e.g., deleting browsing history after closing a browsing session. However, protection against network attackers is in fact provided by Tor. Non-power users are generally unaware of Tor and reluctant to install the Tor browser. As a hybrid, the Brave browser targets privacy- conscious end-users, and, in addition to a private browsing mode, allows users to use Tor-enabled session tabs. We conduct an exploratory online study to investigate users’ perceptions of Brave’s private mode and Tor-enabled sessions to investigate how much Brave’s additional privacy features contribute to (further) misconceptions. We find that Brave’s disclosures did not improve comprehension of privacy and security features; however, Help Center information did. Index Terms—Brave, Private browsing, Web browser privacy, Usable privacy, User study, Misconceptions I. I NTRODUCTION Private browsing is a standard functionality for many browsers such as Chrome, Edge, Firefox, Safari, and Opera. In most browsers’ private modes, browser history is not stored locally and data caching across sessions is prohibited. How- ever, contrary to users’ expectations (which include protections offered by Tor), other privacy and security-related features are usually not offered by default in private mode [1]. Firefox, Opera and a relatively new browser named Brave provide ad- ditional privacy and security features [2]. The Brave Browser offers two private modes: One is comparable to classic private modes, the other offers additional Tor functionality. In an online study with 283 participants, we study user expectations for the three different Brave modes, and how these expecta- tions change when users are shown Brave’s new tab disclosure compares to Brave’s Help Center explanation. We find that Help Center information improves comprehension, while the new tab disclosure does not. II. MAIN OBJECTIVES 1) Considering that general misconceptions about private browsing functionalities are common [3]–[5], we query how end-users understand the different modes of Brave, which is, to the best of our knowledge, not researched. 2) We aim to identify conceptions and misconceptions regarding privacy and security issues for each of Brave’s modes (Standard, Private, Private with Tor, cf. Figure 1). 3) Observation of influence of different disclosures on the comprehension of Brave’s browser mode functionalities. 4) Observation of influence of official information material on the comprehension of Brave’s browser mode func- tionalities. 5) Understand user’s mental models regarding the Brave browser and help to prevent further misconceptions, deriving suggestions about which of the existing in- formation material or disclosure will lead to as most as possible correct conceptions regarding the particular browsing mode. Fig. 1. Disclosures for default, private and private window with Tor mode III. MATERIALS AND METHODS To study conceptions and misconceptions of Brave’s dif- ferent modes, we used the official information material from Brave’s Help Section as well as the official disclosure as stimuli material. Accordingly, eight different conditions were assessed. Within these, we faced our participants with nineteen dif- ferent everyday scenarios adapted from [3], covering general usage, loading speed, and privacy and security features regard- ing availability, visibility, and data retention. For the purpose of our study we cloned the original Brave website (Home, Features, Creators, About, FAQ, and Com- munity) to our own servers and embedded it into an iframe.