Human Factors in Privacy
As mentioned before, privacy, is an even more elusive concept than security. Despite being often clustered with security, privacy interacts with human behavior and desires. At the same time, privacy is also connected with regulations, policy, and operational decisions.
This makes it difficult to understand why privacy issues occur from data alone. For example, in 2023, we published the first paper providing a large-scale perspective on the cloud use of universities. In that paper, we observed remarkable effects, finding a push of universities to cloud infrastructure, all the while also observing differences on country levels. However, from measurement data, we were only able to conjecture as to the root-causes. We now filled this gap with an accepted and soon appearing paper presenting a qualitative study exploring the motivation and reasoning of university decision makers.
In general, explanatory work on privacy is a growing area in the group.
Assessing the Impact of Cookie Consent on Online Behavioral Advertising:
Online behavioral advertising (OBA) is a method within digital advertising that exploits web users’ interests to tailor ads. Its use has raised privacy concerns among researchers, regulators, and the media, emphasizing the need for a reliable mechanism to measure its prevalence. However, there is a lack of systematic research on how user consent choices affect OBA presence, and no open-source frameworks exist for large-scale automated OBA measurement.
To address this, we design and implement OpenOBA, a new framework for automated OBA discovery on the web. OpenOBA is a general, modular, and scalable framework to support essentially any OBA measurement. With it, we conduct a study to measure the impact of three user consent choices for cookies on OBA, uncovering a complex online privacy landscape.
We first confirm the presence of OBA by comparing the increased likelihood of encountering ads from a specific topic, i.e., Style & Fashion, when browsing with an artificially induced behavior versus when browsing without any particular behavior. Then, we find that the Accept All choice significantly raises the number of OBA ads. For the Reject All option, on the other hand, we observe that it reduces the number of unique third-party tracking cookie hosts (tracker domains) by around 70%, yet it still shows ads related to the user’s interests. Notably, we also find that OBA ads are only served through Google-related domains across the three banner interaction configurations used, despite the involvement of up to 191 different tracker domains in the Accept All configuration. This underscores the dominant role of major players in the OBA ad market. Finally, to foster reproducibility and further research, we open-sourced our framework and released all data and analysis scripts.
Cookie Behavior and Partitioning:
The introduction of Cookies Having Independent Partitioned State (CHIPS) marks a significant step toward balancing user privacy with essential web functionalities. CHIPS isolates data within specific contexts, preventing cross-site tracking while maintaining the functionality of websites. However, the adoption of CHIPS in real-world web usage remains largely unexplored.
In this project, we investigate the state of CHIPS deployment, providing an overview of how CHIPS has been integrated into web ecosystems since its introduction. Leveraging the HTTP Archive dataset, we first find that the adoption of partitioned cookies remains slow, with most domains still relying on non-partitioned cookies, though a slight increase in both types is observed starting in early 2024, coinciding with Google’s phase-out of third-party cookies for 1% of users. This sudden onset of the third-party cookie phase-out has resulted in a haphazard way of adoption for some domains, which caused them to overlook important configuration requirements, resulting in improper settings due to limited awareness of the specific guidelines such as SameSite=None and Secure.
In addition, we observe a positive signal for privacy as third-party trackers begin adopting partitioned cookies, with a noticeable increase starting in early 2024. However, as of September 2024, only a small number of trackers have fully transitioned to using partitioned cookies (up to 0.5% of tracking domains), while some continue to rely on both partitioned and non-partitioned cookies (up to 3.1% of tracking domains), highlighting that the shift is still in its early stages, especially for tracking domains. Finally, we observe stark asymmetry among the early adopter tracking domains: some have already added some partitioned cookies to all sites with a presence, while others, notably Google’s doubleclick.com has only deployed partitioned cookies to around 5% of pages where it is present.
Cloud Adoption and Data Privacy:
As universities increasingly rely on digital infrastructures to manage research, administration, and learning environments, two critical challenges emerge: the adoption of cloud-hosted services and the evolving landscape of data privacy under the General Data Protection Regulation (GDPR). While cloud computing offers scalability and efficiency, institutions must navigate complex decision-making processes, shifts in responsibilities, and risks tied to monopolistic markets. Simultaneously, compliance with GDPR requires universities to reassess their privacy policies, balancing transparency, legal obligations, and institutional accountability. This research investigates both dimensions of digital transformation, offering a comprehensive analysis of how higher education institutions navigate technological and regulatory change. We explore cloud adoption in universities through qualitative research, analysing decision making frameworks, stakeholder roles, and governance challenges. Our findings highlight that cloud migration is not merely a technical transition but a shift requiring operational mindset changes. Without such shifts, universities risk replicating inefficiencies that prompted their move from on-premise solutions. In parallel, we examine GDPR’s impact on university privacy policies. Employing a mixed-method approach that combines qualitative content analysis with quantitative textual analysis, we assess how these policies have evolved, focusing on transparency, user rights, and institutional accountability Despite policy expansions post-GDPR, issues such as vague language, fragmented disclosures, accountability evasion and the delegation of compliance burdens to users persist. We provide a comprehensive view of the digital transformation of higher education, revealing the interplay between technological advancements and regulatory adaptation. It emphasizes the need for informed decision-making in cloud migration and greater institutional accountability in privacy practices. By grounding our analysis in empirical evidence, we contribute to a deeper understanding of how universities navigate the intersection of cloud computing, data governance, and regulatory compliance in an increasingly digital world.
Characterizing Online Tracking in Piracy Related Services:
The proliferation of Illegal Movie Streaming Services (IMSS) has posed significant challenges to legitimate streaming services and law enforcement alike, causing financial losses and complicating efforts to combat copyright infringement. Motivated by the absence of a comprehensive list of IMSS, and recognizing that IMSS websites often have short-lived domains, we first introduce a methodology to detect IMSS sites. Our evaluation demonstrates that our method achieves a recall of 84.31% in identifying IMSS. Applying this method on the Tranco Top 1M domains, we find 283 new websites hosting IMSS. When characterizing the IMSS ecosystem, our findings reveal that four specific IMSS sites attract considerable attention, appearing in the Tranco Top 10K domains. Additionally, these sites employ complex redirection patterns, with one site using up to 11 hops to evade detection. Using Google Identifiers, we then uncover 11 cases of co-ownership, where multiple sites share the same identifiers, indicating common operation. Finally, by crawling IMSS sites from seven vantage points (VPs), we investigate online tracking practices on these services—an area that has previously lacked thorough investigation. We find that more than 95% of IMSS include at least one third-party tracker on their websites. Interestingly, tracker presence is lower in the European Union (EU) countries than in other VPs. Furthermore, third-party tracking cookies are not the primary mechanism on IMSS sites; instead, the more invasive and unavoidable fingerprinting techniques are predominantly used for tracking across different VPs.
Understanding Users’ ‘False Sense of Privacy’:
As privacy concerns grow, organizations and policy makers promote the use of privacy-enhancing technologies (PETs) to improve user trust and data-sharing behaviors. However, privacy-enhancing technologies (PETs) are often technologically complex and opaque to lay users. It is challenging to understand and effectively communicate the functionality of complex PETs to the users.
Studies typically assess the impact of new PETs by presenting users with a high-level description of the technology before measuring how this treatment changed their attitude or behavior. These results influence business and regulatory decisions (see Gartner’s Hype Cycle for Emerging Technology).
In this project, recently published at a top-tier venue, we investigate how the simple claim to be using a PET may influence users’ trust into a platform. Using Secure Multi Party Computation (MPC) as an example, we assess whether naming specific PETs and providing generic descriptions impact users’ willingness to put trust in service providers and share their data. Our survey presented three randomized controlled trials with 1,457 participants in a data marketplace scenario. The first group was treated with a PET (MPC), the second group with a fictional PET, and the third with a non-PET, serving as a control group.
Our findings reveal that user trust and data-sharing willingness increased with MPC and the fictional PET, indicating that the high-level description, rather than the technology name, shapes user perception. We conclude that claiming the use of a PET is not an effective method to measure the impact of actually using this technology. However, given their mental model, lay users cannot verify the privacy claims of such descriptions presented in studies or by service providers. This increases the risks of users being deceived into a false sense of privacy, leading them to expose more private data than they otherwise would.