Search

Influencer VPN ads (sponsored segments) on YouTube often disseminate misleading information about both VPNs, and security & privacy more broadly. However, it remains unclear how (or whether) these ads affect users' perceptions and knowledge about VPNs. In this work, we explore the relationship between YouTube VPN ad exposure and users' mental models of VPNs, security, and privacy. We use a novel VPN ad detection model to calculate the ad exposure of 217 participants via their YouTube watch histories, and we develop scales to characterize their mental models in relation to claims commonly made in VPN ads. Through (pre-registered) regression-based analysis, we find that exposure to VPN ads is significantly correlated with familiarity with VPN brands and increased belief in (hyperbolic) threats. While not specific to VPNs, these threats are often discussed in VPN ads. In contrast, although many participants agree with both factual and misleading mental models of VPNs that often appear in ads, we find no significant correlation between exposure to VPN ads and these mental models. These findings suggest that, if VPN ads do impact mental models, then it is predominantly emotional (i.e., threat perceptions) rather than technical., Comment: Accepted for publication at USENIX Security 2025

Wi-Fi-based Positioning Systems (WPSes) are used by modern mobile devices to learn their position using nearby Wi-Fi access points as landmarks. In this work, we show that Apple's WPS can be abused to create a privacy threat on a global scale. We present an attack that allows an unprivileged attacker to amass a worldwide snapshot of Wi-Fi BSSID geolocations in only a matter of days. Our attack makes few assumptions, merely exploiting the fact that there are relatively few dense regions of allocated MAC address space. Applying this technique over the course of a year, we learned the precise locations of over 2 billion BSSIDs around the world. The privacy implications of such massive datasets become more stark when taken longitudinally, allowing the attacker to track devices' movements. While most Wi-Fi access points do not move for long periods of time, many devices -- like compact travel routers -- are specifically designed to be mobile. We present several case studies that demonstrate the types of attacks on privacy that Apple's WPS enables: We track devices moving in and out of war zones (specifically Ukraine and Gaza), the effects of natural disasters (specifically the fires in Maui), and the possibility of targeted individual tracking by proxy -- all by remotely geolocating wireless access points. We provide recommendations to WPS operators and Wi-Fi access point manufacturers to enhance the privacy of hundreds of millions of users worldwide. Finally, we detail our efforts at responsibly disclosing this privacy vulnerability, and outline some mitigations that Apple and Wi-Fi access point manufacturers have implemented both independently and as a result of our work., Comment: Published at IEEE S&P 2024

Since the burgeoning days of IoT, Mirai has been established as the canonical IoT botnet. Not long after the public release of its code, researchers found many Mirai variants compete with one another for many of the same vulnerable hosts. Over time, the myriad Mirai variants evolved to incorporate unique vulnerabilities, defenses, and regional concentrations. In this paper, we ask: have Mirai variants evolved to the point that they are fundamentally distinct? We answer this question by measuring two of the most popular Mirai descendants: Hajime and Mozi. To actively scan both botnets simultaneously, we developed a robust measurement infrastructure, BMS, and ran it for more than eight months. The resulting datasets show that these two popular botnets have diverged in their evolutions from their common ancestor in multiple ways: they have virtually no overlapping IP addresses, they exhibit different behavior to network events such as diurnal rate limiting in China, and more. Collectively, our results show that there is no longer one canonical IoT botnet. We discuss the implications of this finding for researchers and practitioners.

Today's network measurements rely heavily on Internet-wide scanning, employing tools like ZMap that are capable of quickly iterating over the entire IPv4 address space. Unfortunately, IPv6's vast address space poses an existential threat for Internet-wide scans and traditional network measurement techniques. To address this reality, efforts are underway to develop ``hitlists'' of known-active IPv6 addresses to reduce the search space for would-be scanners. As a result, there is an inexorable push for constructing as large and complete a hitlist as possible. This paper asks: what are the potential benefits and harms when IPv6 hitlists grow larger? To answer this question, we obtain the largest IPv6 active-address list to date: 7.9 billion addresses, 898 times larger than the current state-of-the-art hitlist. Although our list is not comprehensive, it is a significant step forward and provides a glimpse into the type of analyses possible with more complete hitlists. We compare our dataset to prior IPv6 hitlists and show both benefits and dangers. The benefits include improved insight into client devices (prior datasets consist primarily of routers), outage detection, IPv6 roll-out, previously unknown aliased networks, and address assignment strategies. The dangers, unfortunately, are severe: we expose widespread instances of addresses that permit user tracking and device geolocation, and a dearth of firewalls in home networks. We discuss ethics and security guidelines to ensure a safe path towards more complete hitlists., Comment: Accepted to ACM SIGCOMM 2023

Since 2006, Turkmenistan has been listed as one of the few Internet enemies by Reporters without Borders due to its extensively censored Internet and strictly regulated information control policies. Existing reports of filtering in Turkmenistan rely on a small number of vantage points or test a small number of websites. Yet, the country's poor Internet adoption rates and small population can make more comprehensive measurement challenging. With a population of only six million people and an Internet penetration rate of only 38%, it is challenging to either recruit in-country volunteers or obtain vantage points to conduct remote network measurements at scale. We present the largest measurement study to date of Turkmenistan's Web censorship. To do so, we developed TMC, which tests the blocking status of millions of domains across the three foundational protocols of the Web (DNS, HTTP, and HTTPS). Importantly, TMC does not require access to vantage points in the country. We apply TMC to 15.5M domains, our results reveal that Turkmenistan censors more than 122K domains, using different blocklists for each protocol. We also reverse-engineer these censored domains, identifying 6K over-blocking rules causing incidental filtering of more than 5.4M domains. Finally, we use Geneva, an open-source censorship evasion tool, to discover five new censorship evasion strategies that can defeat Turkmenistan's censorship at both transport and application layers. We will publicly release both the data collected by TMC and the code for censorship evasion., Comment: To appear in Proceedings of The 2023 ACM Web Conference (WWW 2023)

The popularity of Python has risen rapidly over the past 15 years. It is a major language in some of the most exciting technologies today. This popularity has led to a large ecosystem of third-party packages available via the pip package registry which hosts more than 200,000 packages. These third-party packages can be reused by simply importing the package after installing using package managers like pip. The ease of reuse of third-party software comes with security risks putting millions of users in danger. In this project, we study the ecosystem to analyze this threat. The mature ecosystem of Python has multiple weak spots that we highlight in our project. First, we demonstrate how trivial it is to exploit the Python ecosystem. Then, we systematically analyze dependencies amongst packages, maintainers, and publicly reported security issues. Most attacks are possible only if users install malicious packages. We thus try to analyze and evaluate different methods used by attackers to force incorrect downloads. We quantify your ideas by estimating the potential threat that can be caused by exploiting a popular Python package. We also discuss methods used in the industry to defend against such attacks

Typical security contests focus on breaking or mitigating the impact of buggy systems. We present the Build-it, Break-it, Fix-it (BIBIFI) contest, which aims to assess the ability to securely build software, not just break it. In BIBIFI, teams build specified software with the goal of maximizing correctness, performance, and security. The latter is tested when teams attempt to break other teams' submissions. Winners are chosen from among the best builders and the best breakers. BIBIFI was designed to be open-ended; teams can use any language, tool, process, etc. that they like. As such, contest outcomes shed light on factors that correlate with successfully building secure software and breaking insecure software. We ran three contests involving a total of 156 teams and three different programming problems. Quantitative analysis from these contests found that the most efficient build-it submissions used C/C++, but submissions coded in a statically-type safe language were 11 times less likely to have a security flaw than C/C++ submissions. Break-it teams that were also successful build-it teams were significantly better at finding security bugs., Comment: 35pgs. Extension of arXiv:1606.01881 which was a conference paper previously published in CCS 2016. This is a journal version submitted to TOPS

Typical security contests focus on breaking or mitigating the impact of buggy systems. We present the Build-it Break-it Fix-it BIBIFI contest which aims to assess the ability to securely build software not just break it. In BIBIFI teams build specified software with the goal of maximizing correctness performance and security. The latter is tested when teams attempt to break other teams submissions. Winners are chosen from among the best builders and the best breakers. BIBIFI was designed to be open-ended - teams can use any language tool process etc. that they like. As such contest outcomes shed light on factors that correlate with successfully building secure software and breaking insecure software. During we ran three contests involving a total of teams and two different programming problems. Quantitative analysis from these contests found that the most efficient build-it submissions used CC but submissions coded in a statically-typed language were less likely to have a security flaw build-it teams with diverse programming-language knowledge also produced more secure code. Shorter programs correlated with better scores. Break-it teams that were also build-it teams were significantly better at finding security bugs.