"Perfect Privacy Technology and Chasing Rainbows"

Data-driven innovation, whether in the form of tailored medicine, public services, or efficient industrial production, promises to significantly benefit people and the environment, and provide widespread access to data. However, aggressive data collection and analysis practices raise concerns about societal values and fundamental rights. Therefore, one of the most pressing challenges in unlocking the potential of data-driven technologies is ensuring the confidentiality of sensitive personal data while widening access to the data. A new paper from EPFL's Security and Privacy Engineering Lab (SPRING) in the School of Computer and Communication Sciences contends that the promise that any data use can be solved while maintaining both utility and privacy is akin to chasing rainbows. Assistant Professor Carmela Troncoso, head of the SPRING Lab and co-author of the paper, says there are two traditional approaches to preserving privacy. There is a path that involves using privacy-preserving cryptography, processing data in a decrypted domain, and obtaining a result. The limitation, however, is the need to design highly targeted algorithms rather than simply performing generic computations. The problem with this type of privacy-preserving technology, according to the paper, is that it does not address one of the most pressing issues for practitioners: how to share high-quality individual-level data in a way that preserves privacy while allowing analysts to extract the full value of a dataset in a highly flexible manner. The anonymization of data is the second avenue that attempts to solve this challenge, which involves removing names, locations, and postcodes but the paper argues that the problem is often the data itself. This article continues to discuss new research on why the search for a privacy-preserving data sharing mechanism is failing. 

EPFL reports "Perfect Privacy Technology and Chasing Rainbows"