52nd ACM SIGPLAN Symposium on Principles of Programming Languages (POPL 2025)
ACM Conference on Computer and Communications Security (CCS)
Conference on Empirical Methods in Natural Language Processing (EMNLP)
Conference on Empirical Methods in Natural Language Processing (EMNLP)
Transactions on Machine Learning Research (TMLR)
The common practice of exploitable software which becomes patched, creates a cat-and-mouse game that cannot be tolerated in the presence of critical infrastructure or personal data.
In order to mitigate this cat-and-mouse game, we need new technologies that revolutionize the way systems are build and maintained. Our research area tackles this problem by giving foundations for the system design that incorporate security-by-design and methods for the analysis of existing systems. We currently in particular have a strong focus on conceptually understanding adversarial machine learning and its implications on security-critical systems.
With the advent of Online Social Networks and other Online Services, users, often unknowingly, publicly disseminate tremendous amounts of personal information through their online interactions. All of this information is then readily available to data collectors which use it for personal gain or for malicious actions against the user.
Protection of personal data is therefore of paramount importance in a day and age where data disseminated in the Internet, is completely visible and available to anyone who wants to collect it. In our group we develop foundational methods for quantifying privacy and anonymity in the Internet. Our methods allow for the analysis of existing privacy-enhancing technologies, but also for the development of novel, privacy-enhancing solutions. A particular focus at the moment is privacy assessment and protection for genetic data.
Not only smartphones and tablets have become ubiquitous but also everyday household appliances and infrastructure have been computerized – or became ‘smart’. The endless possibilities of app stores have brought diversity and ingenuity to the way we interact with our world. However, the simplicity of developing and distributing apps together with their omnipresence has made it easy for attackers to gain access to our most personal data or extort us, all under the pretext of being a helpful app.
We conduct research as to how to protect user’s data and privacy on mobile, embedded, and other’smart’ devices, we analyse attacks and data breaches and we construct more secure operating systems.
Usable security and privacy research became an important field of research over the last decade. While many IT security mechanisms offer (very) strong security guarantees in theory, humans are a limiting factor in many cases. Choosing secure passwords, understanding and adhering SSL warning messages or encrypting email is a tough challenge for end users. Developers struggle with using secure cryptographic APIs and webmasters are overwhelmed with configuring X.509 certificates. We collect real data from real users of IT security systems and then build systems to help users make sensible decisions.